The UK government today announced a new commission on FOI. Below are a few quick thoughts, gathered together.
The questions in the remit of the commission boil down to asking ‘is FOI undermining decision-making’ and ‘is it too expensive’? The remit itself is, of course, priming discussion in a particular way, framing it towards two issues of (1) whether FOI is hampering decision-making and (2) whether it ‘costs too much’ . So what does the evidence say?
Is FOI Hampering Decision-making?
Just to put this discussion into context:
- Our 2010 study of FOI in the UK found very few requests for Cabinet documents and also found a broader lack of interest in the decision-making process. Leaks are a far more important cause of openness for these citadels of government decision-making than FOI.
- UK governments since 2005 have used the veto seven (or technically eight) times, compared with 48 times in Australia in the first five years of its own FOI Act. This seems to indicate that ‘dangerous’ requests trying to prise open the very centre of government are relatively few in number, though their psychological effect may be disproportionate.
As I’ve said before, the effect of FOI on policy discussions generates lots of heat but very little evidence. Tony Blair claimed FOI had led to more caution over recording decisions or inhibitions in discussion (the so-called ‘chilling effect’). Former Cabinet Sectary Gus O’Donnell also claimed it has ‘hamstrung’ government, though when pressed he could only offer isolated examples-one hypothetical and one based on the coalition negotiations, one of the most unique and unusual political events in recent decades. Doubtless we’ll hear similar claims made again.
In terms of harder evidence, the Justice Committee ‘was not able to conclude, with any certainty, that a chilling effect has resulted from the FOI Act’ and also felt the protections for policy were sufficient and was ‘cautious about restricting the rights conferred in the Act in the absence of more substantial evidence’. The committee argued against change but cautioned care…
Given the uncertainty of the evidence we do not recommend any major diminution of the openness created by the Freedom of Information Act, but, given the clear intention of Parliament in passing the legislation that it should allow a “safe space” for policy formation and Cabinet discussion, we remind everyone involved in both using and determining that space that the Act was intended to protect high-level policy discussions. We also recognise that the realities of Government mean that the ministerial veto will have to be used from time to time to protect that space.
Our own studies found a few examples but no systematic behavior changes around advice or space-and also many officials more concerned about the dangers of not having a record if a judge came knocking.
But the claim won’t go away. The recent Supreme Court ruling weakened the government’s veto over requests and has undoubtedly caused concern, or at least reignited old worries. But is also partly psychological. Politicians believe it happens and keep repeating it, so it then becomes true to them. It is a rather wonderful example of a self-confirming myth, especially as the myth itself may then make people wary. It is, of course, more politically, a convenient and half acceptable way of attacking FOI, for those politicians who don’t like the disruption FOI brings.
It ‘costs too much’
This is a tricky one. Measuring the cost of FOI in any reliable way is almost impossible. Estimates have varied from
- A pre-Act estimate of £350 per request
- A UK government study of 2006 giving £293 per request
- A Scottish government of 2010 of £193 per request.
- Cornwall Council calculated an average of £150 per request
- Bexley council found it was £36 with most requests costing around £19.
As we pointed out here on pg. 33:
- The costs of FOI are very difficult to measure and calculate. Different studies have used different methods and, unsurprisingly, have come to very different results. While you can simply multiply hours by time taken this may fail to catch, for example, the ‘opportunity costs’ of involving other staff or time spent in discussions. By contrast, the 2006 UK study factored in ministerial time (which is, of course, pretty expensive) while the Scottish study included one request that had taken 200 hours-a little over the 18 hours allowed under legislation. See this report.
- The cost of FOI is a political issue. From the view point of politicians and officials FOI introduces ‘concentrated costs and dispersed benefits’ (Fung et al 2007: 117). It is easy to see the resource and, for politicians, the political costs but much more difficult to quantify or see the benefits flowing from FOI, such as transparency. This means there is a hidden bias in any discussion of FOI which tilts discussion in a negative direction.
The real difficulty is arguing in favour of rather vague (but real) ‘democratic benefits’ against concrete numbers. Yet on a day when an FOI revealed that the UK government has been involved in military action in Syria, despite a Parliamentary vote against it in 2013 the argument is there-and from potholes to extraordinary rendition there is a strong case to be made. Not to say that FOI is perfect or unabused but it is an important, if messy, democratic force.
Neither of the two ‘problems’ the commission is looking into are new. In fact, they represent two of the most frequent complaints or lines of attack on the legislation from Ireland to India. Tony Blair, as we all know, felt FOI was ‘abused’ and was ‘utterly undermining of sensible government’ while Cameron spoke of how some requests were ‘furring up the arteries’, with the Act as a sort of cholesterol on the healthy body politic. The issue of resource costs is, of course, perennial in FOI with police forces and councils keen to complain of frivolous costs. Moreover, FOI discussions are often about what those in power think is happening or what they want to believe, the myths, perceptions or rumours rather than the reality.
The interesting point about the remit is that it tilts all discussion naturally towards the two issues of damage and costs, rather than any more equal cost-benefit analysis. How easy will it be to say ‘actually, it doesn’t…’?