A ground breaking study advocates the creation of a new register to prevent policy research from going missing - especially if its findings are inconvenient
As so often, the verdict is cock-up not conspiracy. A ground breaking inquiry into the publication of research commissioned to support public policy shows that much of the fruits of this work - which costs £2.5bn a year - are never shared across government, let alone with the public.
Missing Evidence, carried out by the pressure group Sense About Science, also confirms that findings which confirm ministers’ gut feelings are more likely to be published promptly and with fanfare than those that do not. Examples include studies on the use of food banks and international comparisons of drugs laws.
However the inquiry’s author, former Court of Appeal judge Sir Stephen Sedley, concludes that the reason so much expensively acquired knowledge disappears into a black hole is not censorship but the lack of arrangements for sharing it. His proposed solution: to create a standardised central register of all externally commissioned government research.
The study finds ample evidence of inadequacies in current practice. “There is good evidence of pressures to align publication of research both with government's broad policy agendas and with particular policies relating to the evidence itself,” Sedley says. Sometimes the reasons appear absurdly trivial: a report on animal research was delayed because the prime minister’s special advisers thought it clashed with Alzheimer’s Week.
But it also comes up with a less predictable finding, that Whitehall simply does not know what is going on.
The inquiry submitted freedom of information requests to 24 ministerial departments and two executive agencies asking for full lists of studies they had carried out themselves or commissioned from outside experts between 2013 and 2015. Almost incredibly, 10 departments said that finding the information would be too time consuming and costly. Among these were the Treasury, the Cabinet Office and the Home Office.
But a handful, notably the Department for Environment, Food and Rural Affairs (Defra), have a register with all research tracked in public view.
Sedley said the discovery that many departments either do not possess or cannot easily provide such basic information gives his report “a new and unexpected dimension”.
Quite apart from fulfilling the public’s right to know, publishing research in an accessible way would save the government the time it spends hunting for material as well as enabling departments to learn from each other’s work.
To fears that a blanket call for all government commissioned research to be openly published might make policy-makers less likely to commission research, Sedley says this does not appear to have been the case at Defra.
However, the report suggests that existing web systems are not up to the job. It quotes one respondent saying “it would never occcur to me to try to find research on gov.uk”.
To fill this gap, Sedley recommends “building on existing departmental research databases” to create “a searchable list of commissioned external research”. Each study would be listed under a unique identifier and be linked to associated datasets and comprehensive metadata.
This approach makes sense; certainly more so than trying to set up a discrete research database. Older readers may remember a system called the Knowledge Network which the Cabinet Office installed at the turn of the millennium. The Lotus Notes-based software from IBM was supposed to introduce online collaborative working across departmental boundaries by allowing them to share information and knowledge. It failed, mainly because the Cabinet Office lacked the clout to force departments to share.
Some cultures have changed since then, particularly when it comes to data. The proposed new system would be based on open data standard, Sedley says. More important, it must be underpinned by “a clear statement of the current requirements for prompt publication and adherence to them”. Enforcing that will no doubt remain the central issue.