Justice CommitteeWritten evidence from Greg Muttitt

Executive Summary

1. The Freedom of Information Act 2000 (FOIA) has become an important element of our democracy, and is working well in making government more accountable.

2. However there are four respects in which FOIA is not working effectively:

I have experienced considerable delays, primarily in the conduct of public interest balancing tests and internal reviews. There is a case for amending FOIA to introduce time limits in these areas.

There remains a culture of secrecy within public authorities, manifested in foot-dragging or evasion with many requests. The Information Commissioner’s Office (ICO) and Information Rights Tribunal (IRT) therefore play a vital role in delivering the public’s rights to information.

There is an inherent structural flaw in FOIA, in that a reticent public authority has no incentive to release information, and every incentive to refuse, as it can always release at a later date if forced to do so. The biggest potential shift to the culture and practice of government openness would be brought about by meaningful penalties for public authorities that violate FOIA.

Public authorities routinely fail or refuse to give meaningful advice or assistance under §16 of FOIA, offering instead unhelpful generalisations.

3. The UK’s FOIA has two particular strengths:

That complaints and appeals are heard by the ICO and IRT rather than by courts makes the process accessible to those who do not have access to lawyers.

The consideration of the balance of public interest in relation to exemptions emphasises the release of information which is of value to the public.

4. There are two inherent weaknesses in FOIA:

The class-based exemptions of security-related information and the exclusion of security bodies from the Act’s scope.

The §53 veto.

5. There are four further weakness in how FOIA is applied in practice:

Public authorities often apply unintelligent methods of searching, resulting in information not being found or conversely in requests being unnecessarily pushed beyond the §12 limit.

Public authorities do not give specific reasoning for rejecting requests, resulting in unnecessary complaints to the ICO.

The concept of public interest is often interpreted in an unbalanced way, as broad reasons for maintaining an exemption are compared with narrow reasons for releasing information.

It is unhelpful and unnecessary under §40 to exempt the names of those acting in a purely professional capacity.

6. The calls by certain public authorities for reducing the §12 cost limit are misplaced, as other solutions could better align minimising public costs with the public interest in government accountability, which is indeed the original purpose of the Act.

7. It would also be unfair, and would hamper FOIA, to include reviewing and redaction in the §12 cost assessment, since the amount of time spent on those activities is decided by the public authority rather than the requester. A better solution would be to charge requesters per page for copying and sending information, preferably with clear information as to what information is held.


8. I am an experienced requester. Since FOIA came into force in 2005, I have made over 70 requests, mainly to central government departments. I have also made around 50 requests under the US Freedom of Information Act, which gives me a comparative perspective.

9. Many of my requests were made in the course of researching my book, Fuel on the Fire—Oil and Politics in Occupied Iraq, which was published by Random House in April 2011. I am also Director of Campaigns and Policy at the charity War on Want. In this submission I draw on experience from both roles.

Does the Freedom of Information Act (FOIA) Work Effectively?

10. Broadly, the Act works effectively. For example, through FOIA I have obtained and published significant information revealing government mistakes and misleading statements in respect of the Iraq War. FOIA has served a profound public interest in helping the public to learn the lessons of that War, and in disclosing government actions.

11. Information I have obtained through FOIA has generated many press stories, including on the front page (Independent, “Secret memos expose link between oil firms and invasion of Iraq”, 19/4/11).

12. At War on Want, information obtained through FOIA has enabled the charity to strengthen its arguments and target its campaigns, enhancing its ability to achieve its charitable aim of relieving poverty. For example, in respect of our campaign to protect the livelihoods of crop growers, we obtained records of the supermarkets’ lobbying of the government against regulation, allowing us to hone our arguments.


13. A key aspect in which FOIA’s effectiveness is limited is in excessive delays by public authorities, especially in conducting internal reviews. For example, one Cabinet Office review of my request took a whole year; a Department for Business, Innovation and Skills (BIS) review took 21 months—even after repeated interventions by myself and by the ICO. Whilst these are extreme cases, I have frequently found internal reviews to take more than the ICO’s recommended maximum 40 days, and usually for no good reason: internal reviews have mostly failed to show any serious reconsideration of the issues.

14. RECOMMENDATION: The Code of Practice and/or FOIA should be amended to set a time limit of 20 working days, or forty days in exceptional cases, for the conduct of internal reviews.

Lingering Culture of Secrecy

15. Comments in the press—most notably by former Prime Minister Tony Blair and former Cabinet Secretary Gus O’Donnell—suggest that some want to return to the days when government had no accountability to the citizenry. My experience of using FOIA has often found recalcitrance by public officials, including a default position of rejecting requests, failure to substantively consider arguments and holding back information until the last minute when forced to release it by the ICO or IRT.

Lack of Incentive to Comply

16. A major impairment to FOIA’s effectiveness, in the context of civil servants’ frequent reluctance to release information, is that a public authority faces no penalty for wrongfully refusing requests. An authority can routinely refuse all requests, on the assumption that the majority of requesters will eventually give up. In the small percentage of cases that are pursued, the authority can release information at the later stage. From the perspective of a recalcitrant civil servant, such a strategy would have every advantage and no disadvantage.

17. Indeed, my experience is that authorities frequently seem to reject requests as a default position, and only to provide information when made to do so by the ICO or IRT. To illustrate, during 2009 and 2010 I made 38 requests. Of the 27 cases where the information was held, in only nine cases was the information released at the first stage and in one case after internal review. In 12 cases the information was released only after the ICO either issued a Decision Notice or intervened informally and in one case after an IRT ruling (four cases were exempt).

18. This suspicion of default rejection is reinforced by events on the two occasions I have had cases at the IRT. In both cases the public authorities (respectively BIS and the Cabinet Office) insisted information was exempt right up to the point where the IRT process was under way, at which stage they started releasing information, apparently in order to avoid an adverse ruling.

19. If the government is genuinely concerned about the burden of FOIA on its resources, a sensible route to reducing the costs would be to properly consider and respond to requests in the first place, rather than rejecting requests and so having to put greater resources into an ICO investigation or an IRT hearing (including the high costs of lawyers).

20. RECOMMENDATION: public authorities should be fined by the ICO and IRT when they have handled a request incorrectly, with the level of fine set so as to have a deterrent effect, and varied according to the degree of negligence or bad faith shown by the public authority.

Failure to Give Advice and Assistance§16

21. In almost all cases when I have requested advice under §16, I have been given a non-specific, largely meaningless response based on generalities. As a result, my requests have not been as well honed as they might have been, wasting my time and that of the public authority. RECOMMENDATION: Again, the ICO and/or IRT should have a power to fine public authorities.

22. A further problem is long delays in the provision of advice: for example one of my §16 requests finally received the advice four months later, after I had chased it five times. RECOMMENDATION: §16 should be amended to set a time limit of ten working days for the provision of advice.

The Strengths of FOIA

Non-judicial mechanism

23. One of the strengths of the UK’s FOIA compared to the US equivalent is that the appeal and complaint are heard by non-judicial bodies, namely the ICO and IRT. This allows information to be obtained by those without the resources to hire lawyers. For example, I recently represented myself in a case at the IRT and won.

24. Both the ICO and IRT are functioning well. The ICO’s guidance notes are especially appreciated.

Concept of public interest—sections 2(1)(b) and 2(2)(b)

25. Another strength is that most of the exemptions are qualified by the public interest test. This introduces a discretionary aspect, which is important because legislation is not capable of anticipating all circumstances that may arise. It also places the Act clearly in the service of the public interest.

The Inherent Weaknesses of FOIA

National security—sections 23 and 24

26. The exclusion of certain security bodies from the scope of the Act and the exemption of information obtained from them are unwelcome departures from the general principle of discretion based on the balance of public interest. The same is true of the criterion of ministerial certificate in exempting information whose release might damage national security. “National security” is a concept over-used by governments to justify secrecy, and must always be balanced against the public interest.

27. While the public interest threshold would be set high for security-related information, recent revelations of security services’ involvement in torture and extraordinary rendition demonstrate the importance of also those services’ accountability. For comparison, in the USA’s FOIA does not exempt the CIA.

§53 veto

28. Although the §53 veto has only been used twice, there seems no justification for giving the ultimate say to the government itself. This provision undermines the credibility of and confidence in FOIA.

Weaknesses in how FOIA is Applied in Practice

Unintelligent searching

29. Public authorities often search for requested information in a manner that defies common sense. The effect is either to fail to find relevant information or conversely to find so much that it cannot be processed within the §12 limit. Many of my requests have been for information that would be routinely required in the course of the authority’s usual business, yet have been met with a response that the search is impossible to conduct.

30. For example, a recent request by War on Want to UK Trade and Investment for records of UKTI’s recent communications with four named companies was rejected under §12, on grounds that to find the information “we would need to consult every member of UKTI throughout the world”.

Failure of public authorities to give reasoning, even at internal review

31. When information I have requested has been considered exempt, public authorities have invariably given generic descriptions of the relevant exemption, rather than any reasons why they apply in the particular case.

32. In some cases, clear reasoning, especially at internal review stage, might have shown me that the authority was correct, in which case I would not have complained. More often though, I have ultimately found that the authority has no reason for applying exemptions, and only even considered the request specifically when contacted by the ICO. This has led to considerable wasted resources: mine, the ICO’s and the government’s.

Public interest interpreted in unbalanced way

33. A general problem with the public interest balancing exercise has been that authorities have not compared like with like. The public interest in maintaining an exemption has been taken to be broad, in that release would affect a whole class of information, but the public interest in release has been taken to be narrow, relating only to the specific case. The ICO has also committed this error on several occasions.

34. The exemptions that have come up most commonly in response to my requests are those under §§27, 40 and 43. In relation to §§27 and 43, it has been argued that release of information would discourage foreign states or companies (respectively) from sharing information with the government in future—a general public interest. Conversely, it has been argued that releasing the information would serve the public interest in respect only of its specific contents, not in terms of a broader public interest in transparency and accountability. For an example of this, see ICO Decision Notice FS50286465, 15 December 2010.

Names of individuals—§40

35. The redaction of people’s names often diminishes the value of information. The release of what someone has said or done in a professional capacity should not in general be considered unfair. Case law has however established an approach where names are redacted of all officials below Senior Civil Servant grade. There should be no reasonable expectation of privacy of a person’s actions carried out in a professional capacity, as such information is routinely available within professional circles, such as through industry conferences. RECOMMENDATION: §40 should be amended to such that names of individuals acting in a purely professional capacity are not treated as “personal data”.

Costs and the Burden on FOIA on Public Authorities’ Resources§12

36. Some public authorities have complained about the cost of responding to information requests. Sometimes these calls have drawn attention to specific examples of frivolous use of FOIA: such as requests for the number of toilet rolls or teabags purchased by an authority, or requests for information on alien abductions. It would make sense to charge commercial organisations who use the Act to help them sell goods or services to public authorities. However the cost limit of public interest cases should not be reduced, as that would undermine FOIA’s ability to deliver the government accountability it was intended for.

37. In particular, media requests—interpreted broadly to include bloggers, authors and campaigners—should not be restricted, given the important role media play in holding government to account.

38. Statistics from Frontier Economics suggesting that 5% of requests account for 45% of costs may create a temptation to exclude the most complex and costly requests from the Act by lowering the §12 limit. This would be unwise as often it is the more complex requests that give us the greatest insight into public authorities, and which best expose bad practice.

39. It would also be wrong to include the time spent reviewing and redacting in the assessment of whether a request exceeds the limit—because the time spent on those activities is chosen by the public authority not the requester. If the Fees Regulations were changed to include review and redaction costs in respect of §12, reticent public authorities would have an incentive to be excessively rigorous in their reviews, thereby automatically pushing many requests unjustly above the cost limit.

40. On the other hand, the current arrangement does create perverse incentives for requesters, whereby a request is less likely to be rejected under §12 if it is framed more broadly—as that will reduce the time spent locating and extracting the requested information—at the expense of requiring more time to read, review and redact. The result is that sometimes an authority will spend time processing information that is in fact of little interest to the requester.

41. A good solution to this problem would be to add another stage to the FOIA process, wherein an authority would first provide a requester with a list of documents in scope of the request, including title, date, metadata and number of pages. Charging copying costs (as provided in the Fees Regulations) would discourage requesters from asking for irrelevant information, without excessively deterring requests as a whole. The list of in-scope documents would then effectively serve as a price list. A selection and payment process could be managed online, leading to a second stage of processing which included review and redaction, focused only on the documents the requester actually wants.

42. In general, the cost to public authorities should be better aligned with the purpose of FOIA: by rejecting or charging for requests that do not improve government accountability. Such decisions should still be subject to complaint and appeal to prevent abuse.

43. The USA’s FOIA has a sensible approach to this problem. Requesters are divided into categories, and charges levied accordingly:

Commercial requesters are charged the full costs of search and review (ie staff costs) as well as of copying and sending.

Individual requesters are charged for copying and sending costs, and for any search and review time beyond two hours.

Media, educational and non-commercial scientific requesters are charged for copying and sending costs but not search and review time.

44. There is then a waiver of charges for requests in the public interest, defined as those where the information is likely to contribute to public understanding of government activities and is not in the commercial interest of the requester.


45. In conclusion, the passing of FOIA has been a major achievement in enhancing the accountability of government. Post-legislative scrutiny of FOIA creates an important opportunity to correct some of the specific weaknesses of the Act. However, calls to restrict the application of FOIA—whether on cost or other grounds—should be resisted, as they would undermine that great achievement.

February 2012

Prepared 25th July 2012