The Campaign’s views on the operation of the Open Government Code of Practice

NOTE: The Code of Practice on Access to Govenment Information was introduced by the Conservative Government in April 1994 as an alternative to a freedom of information act. The Code has now been superseded by the Freedom of Information Act 2000 which came fully into force on 1 January 2005. An archived version of the Code is available on the The National Archives website here.

Evidence to the Select Committee on the Parliamentary Commissioner for Administration (the Ombudsman) on the operation of the Code of Practice on Access to Government Information

March 16, 1995

The Campaign regards the Code of Practice on Access to Government Information as a significant advance, though one that has substantial weaknesses. We believe that a Freedom of Information Act, creating a statutory right of access to information across the whole public sector, is needed.

On the positive side, the code:

1. limits the grounds for withholding information

2. provides, for the first time, a mechanism for the independent review of refusals of information by officials and ministers

3. introduces the important principle that even exempt information may be disclosed if there is an overriding public interest in openness

4. commits departments to releasing the internal guidance they use in their dealings with the public.

But the code’s drawbacks are that:

1. it offers only information, not copies of actual documents – a fundamental weakness, though one that the Parliamentary Ombudsman has challenged

2. its scope is limited to bodies subject to the Parliamentary Ombudsman’s jurisdiction. Although separate codes for the NHS and local government have been promised, important areas of the public sector fall outside the scope of any code

3. some of the code’s exemptions are excessively broad, protecting information which in our view should be disclosed

4. the arrangements for charging permit high fees to be charged, with no provision for waiving them for non-commercial requesters or information whose disclosure would be in the public interest

5. because the code has no legal basis it cannot override the 250 or so separate statutory restrictions on disclosure which may prevent the release of many kinds of information

6. although there is considerable pressure on departments to comply with the Ombudsman’s recommendations, at the end of the day departments remain free to reject them.

Three of these shortcomings can be remedied only with legislation. This would be required (a) to extend the code to bodies not currently subject to the jurisdiction of the Parliamentary Ombudsman (b) to override existing statutory secrecy restrictions and (c) to ensure that departments and authorities must comply with any disclosure scheme.

The Campaign has been using the code itself; and encouraging others to do so. In May 1994 we published a detailed briefing on the code which has been circulated to the Campaign’s supporting organisations and others. We think it important that people know of their rights to information and be encouraged to use them. The public’s assumption that official information will not be disclosed is so prevalent that, unless the code is widely publicised, people will continue to assume that there is no point in asking for information.

Our preliminary experience of the code in operation suggests that it is capable of eliciting information which would previously not have been disclosed. The prospect of an investigation by the Ombudsman does appear to make departments whose objections to disclosure are not well-founded think again. The release of internal guidelines is likely to be valuable to individuals in their dealings with government and to organisations which advise them.

The increased funding to the Ombudsman’s office, to enable him to take on new staff to deal with code complaints, is welcome. So is the publication of the Office of Public Service and Science’s (OPSS) substantial internal guidance on the code. Although we think the arrangements for monitoring the handling of requests under the code should be improved, the fact that systematic monitoring across departments is taking place is a further positive development.

However, the code has not been adequately publicised. Rights which people have, without knowing of them, are of limited value. Requests under the code have sometimes been met with prohibitively high fees. Some requests have been dealt with in an unhelpful manner, with departments citing multiple, and sometimes clearly inappropriate, exemptions. There may be excessive delays both in replying to requests, and in carrying out the internal review required before a complaint to the Ombudsman can be made. In some cases the series of obstacles between the dissatisfied applicant and the Ombudsman may be so great as to effectively deny that remedy to requesters. Ministers have claimed that the code’s advantage over a Freedom of Information Act is that it is more informal and less bureaucratic than legislation. But there are signs that some of the arrangements under the code are themselves legalistic and bureaucratic, without the benefits of a legally enforceable right.

Level of code requests

It is clear from the Parliamentary Ombudsman’s report on the first eight months of the code [HC 91] that few complaints to him are being made.

The OPSS report on the code’s first 8 months states that a total 2,600 requests under the code have been made. We wonder whether this figure may be an over-estimate. The Welsh Office alone accounts for 1059 requests, representing 40% of the total, a surprisingly large proportion.

The Welsh Office requests do not appear to comply with the definition of a “code request” set out on page 3 of the OPSS report. Instead they are described as requests ” which are outside the run of information usually asked for, whether or not the applicant specifically mentions the Code”. There is a great discrepancy between the Welsh Office figure, and the modest number of requests received by the Scottish Office (which received 45 requests) and the Northern Ireland Office and departments (which between them received only 5 requests). This may be because the Welsh Office has been more successful in publicising the code. However, it may be significant that not one out of more than a thousand requests was refused, even in part. One would normally expect at significant proportion of applicants to be seeking information which in fact was legitimately exempt. Customs and Excise, for example, refused 4 out of the 20 requests; 3/12 MAFF requests were refused in part; 3/9 Foreign Office requests were partly refused. It is hard to believe that out of more than a thousand similar requests to the Welsh Office not one happened to be seeking even a fragment of exempt information, and that all 1059 could have been met full. This suggests that the Welsh Office may be including more straightforward requests which most of government does not regard as made under the code. The fact that some 300 code requests are said to have come from the media – an extremely high figure – raises the question of whether some normal press requests are also being classified as made under the code.

Definition of a “code request”

We suggest that the OPSS definition of a “code request”, in paragraph 8 of its 1994 report, might usefully be amended to ensure that applicants whose requests are dealt with under the code are always told that this has been done, even if they have not themselves cited the code.

The department’s response to a request should therefore (a) always state that it was dealt with in accordance with the code; (b) if information has been withheld under an exemption, acknowledge the fact; (c) refer to the code’s appeal process whether or not information has been withheld; this will alert requesters who may feel that a reply is not entirely adequate to the fact that there is something they can do about it; (d) enclose a copy of the department’s own code leaflet or written procedures.

This procedure would help both to publicise the code, and expose the department’s responses to the disciplines of the code. Once this had been done, the request might then legitimately be recorded as a “code request”.

Publicising the code

A number of factors may partly explain the low use so far made of the code.

, the government gave little publicity to the code’s launch. The fact that the code was brought into force on a bank holiday during the parliamentary recess, guaranteed an inconspicuous start. Although the draft code was published in the July 1993, the final code was published only three working days before its implementation on April 4 1994. Journalists who might otherwise have written stories anticipating the code’s introduction had no final text until the last moment, and little opportunity to publicise the event.

Second because the code has been implemented administratively and not – as would have been the case with a Freedom of Information Act – gone through the legislative process. As a result the public awareness and expectation that would have been created by a bill going through, and perhaps being fought over in, succeeding parliamentary stages has not occurred.

Third, many of those who have heard of the code may have been disinclined to use it because of its shortcomings. For example, the fact that the code – at least as interpreted by the government – commits departments to releasing only a letter with summary information, rather than copies of documents, may suggest to applicants that it offers relatively little that is new. People have always been able to seek information in correspondence with departments. Moreover, those seeking access to their own personal files – perhaps the largest group of potential users – will have been specifically deterred by the explicit statement that “the Code should not be regarded as a means of access to original documents or personal files” [Code, paragraph 8].

Fourth, the government has spent little on publicising the code. The OPSS and many individual departments, have produced leaflets explaining the code. Nevertheless, the government’s total expenditure on publicity for the code during 1994 is said to be £51,157 [Written Answer, 20.12.94. col 1013]. Some of the individual figures which make up this total appear to be overestimates. For example, the Office of Fair Trading is said to have spent £6,836 publicising the code, a surprisingly high proportion of the total. In fact the OFT has told us that this figure refers to the cost of printing a general booklet on the OFT’s charter responsibilities. Only part of this can be attributed to the code, as the booklet also deals with the OFT’s standards of service, complaints procedures and statutory functions. The Northern Ireland Office is said to have spent £2,945 publicising the code. The NIO has told us that most of this sum was spent on internal circulars informing staff of the code, and that only £830 of the total was spent on information for the public.

The publicity budget of £51,000 for the code’s first eight months may be compared to:

* the £542,010 spent publicising the Charterline during its first 9 months [WA, 2.2.94, col.789];

* the £311,747 spent on publicity for the 1993 Chartermark winners [WA 24.3.94, col 358];

* the £420,000 publicising the citizen’s charter first report and producing summary guides to it [WA 18.3.93, col 342].

* the initial £291,550 spent distributing and publicising the council tenant’s charter [WA 27.1.92, col 415]

* the nearly £2 million spent circulating copies of the parents charter to all schools in September 1991 [WA 14.7.92, col 556]

* the £2 million spent producing and distributing the patient’s charter up until January 1992 [WA 21.1.92, col 151]

There has been no paid advertising for the code. By contrast, we note that in 1993-94 the Department of Employment spent over £1.8 million advertising the existence of career development loans [WA, 18.1.95, col 568]. A more relevant example may be found in the budget for publicising the Data Protection Act, which amongst other things gives individuals the right to see computerised personal files held on themselves. In recent years the Data Protection Registrar’s annual budget for publicity – which is directed both at informing individuals of their rights and computer users of their obligations – has varied from around £400,000 pa to over £750,000 pa. Following a television advertising campaign in 1993-94, a survey showed that 64% of people had heard of the Data Protection Act or Registrar – a 10% increase on the previous year [10th Annual Report, pages 54 and 83].

We have no doubt that people will use new rights to information if they know of them and expect them to be effective. Recently, new Environmental Information Regulations came into force, implementing a European directive, and creating a broad right of access to environmental information. Most departments do not record the numbers of requests made under them. However, one that has done so, the Northern Ireland Office, stated that it had received 1,221 requests in the eight months, April to December 1993, after the regulations came into force in Northern Ireland [WA 6.7.94, col 253-4]. The contrast between this figure, and the total of just 5 requests to the Northern Ireland Office and departments in the first 8 months of the code’s existence is striking. We suspect it indicates a lack of awareness of the code’s existence – not that the people of Northern Ireland are concerned solely with environmental issues to the exclusion of all other aspects of government activity.


One of the code’s most significant defects is that it promises access only to information, not to documents:

“There is no commitment that pre-existing documents, as distinct from information, will be made available in response to requests”. [Code, paragraph 4]

The OPSS guidance on the code states:

“The Code…does not constitute a right of access to documents or records.” [Part I, paragraph 50]

“Requests for information should be met in the simplest, most cost-effective way. This will usually be by letter giving the information requested. The public has no right, under the Code, to a copy of a particular document.” [paragraph 52]

“A digest of the relevant information is likely to be more useful to most applicants than reams of less relevant original material.” [Open Government, Summary of Consultation Comments & Government Response, March 1994]

In our view this is a potentially overwhelming defect: the opportunities for selective editing are obvious. A body which anticipates that full disclosure may expose it to criticism is unlikely to be able to resist the temptation to produce a summary which slants or even conceals the truth. There is little reason for the public to have confidence in such a disclosure scheme.

We do not accept the suggestion that the public will prefer a specially prepared “digest” of the “relevant” information to the original documentation. In our experience, both under the code and other disclosure schemes, people often seek information on subjects about which they have considerable knowledge and are capable of making use of whatever “raw data” may exist or have access to experts who can do so for them.

The former Minister for Public Service and Science, Mr William Waldegrave, also argued that to release a specially prepared summary “will be cheaper in most cases” than to supply the original documents. [“Open Government” seminar, 21.10.93]

However, we think it likely that preparing a balanced and accurate summary from a variety of disparate and perhaps inconclusive documents in a file will involve a great deal more painstaking work, and be considerably more expensive, than releasing the documents themselves.

The OPSS guidance appears to acknowledge that this will sometimes be the case. Nevertheless it suggests that summaries should still be disclosed, even when releasing copies of originals would be cheaper:

“Charges will not be made to cover….excessive time spent on the preparation of new material when it would have been practicable within the policy set out in the Code, and more cost effective, to provide edited copies of a pre-existing document” [my emphasis] [Part III, paragraph 14]

The code’s approach is also said to be preferable because:

“People will in general find it easier to describe the information they seek, rather than the documents they wish to see.” [Open Government White Paper, paragraph 4.8]

If the guiding principle is, as this appears to suggest, the convenience of the requester, then the most satisfactory solution would be to allow applicants to present requests in their own way – not to exclude requests for documents. Someone who wants an answer to a particular question may need merely to put the question; but those who feel they need to see an existing document should be able to ask for it.

At present, the code implies that if a requester becomes aware of an unpublished report which may be of interest he or she cannot ask for it but must specify the particular information from it that is required. The limitations of this approach are self-evident: without seeing the document, the applicant will probably not be able to describe its contents. Even if some of its contents can be guessed, there may be other more relevant information in it whose significance cannot be predicted, and which the applicant may fail to ask for. We see no reason why access should be limited in this artificial way. The straightforward solution is for people to be able to ask for and obtain a report which they may ask for, or documents containing information of a kind they have described, subject to the deletion of any exempt information.

It is notable that a similar code of practice which has just been introduced in Hong Kong, rejects the government’s approach to documents. Hong Kong’s ‘Code on Access to Information’ which came into force on March 1 1995 is in most respects identical to the UK code. But it states:

“Where a request, whether written or oral, cannot adequately be met by an oral reply or provision of a standard leaflet, form, etc., the information may be given by –

* providing a copy of the relevant record or part thereof

* providing a transcript of the relevant record or part thereof

* affording a reasonable opportunity to inspect, hear or view the relevant record or part thereof, or

* providing a summary of the relevant record or part thereof.

So far as possible information will be provided in the form in which it exists. Where disclosure of certain information in a record is to be refused, access will normally be provided to the remaining part of the record.”

The official guidance on the Hong Kong code adds that it is “preferable” to provide a copy of the original record, and that provision of a summary should be considered only if a document contains so much exempt information that a copy from which the exempt passages have been deleted would be “meaningless or misleading”.

We are pleased to see that the Ombudsman has, in effect, rejected the government’s approach to documents. He has stated:

“One aspect of the Code which has been criticised by some is that it puts no obligation on a department to let the public have access to documents as opposed to information …My remit is to investigate and report whether bodies within my jurisdiction have complied with the requirements of the Code. I would not therefore criticise a body if it had fulfilled its obligations under the code without releasing copies of documents involved. However I normally construe a request for documents as meaning that a complainant is seeking all the information contained in the document specified and, save where all or part of that information can legitimately be withheld under the exemptions contained in Part II of the Code, I normally expect all that information to be released. Thus, while there may be exceptions, there are likely to be a number of occasions when…I conclude that the most practical way to release the information sought is to provide a copy of the actual document in which that information is contained.” [HC 91]

This is an impressive attempt to overcome the code’s defective approach to documents. However, unless the government accepts the implications of the Ombudsman’s statement, it may not resolve the problem. The Ombudsman may urge a department to release a document – yet not be able to formally criticise it should it refuse to do so. In other cases, applicants who do not have the stamina to persist in taking a complaint through all the required stages up to the Ombudsman, may be dealt with by a department on the basis of the code and OPSS guidance and receive a precis rather than the full information. In our view, the government should now formally amend the code so as to require the provision of copies of original documents where these are asked for or where this would be the most complete way of providing the requested information.

Delay in responding to requests

The code sets out a target of 20 working days for responding to simple requests. The majority of requests from the Campaign appear to have received a response within this period.

However, the target was substantially exceeded in a number of cases:

* A straightforward request to the Department for Education took 34 working days. The request did not require the collation of information, and was eventually dealt with by forwarding to the Campaign a copy of a 2-sided summary of market research which had been previously provided to the DfE by the Central Office of Information.

* A series of related requests to departments, seeking information on the information they may have received from British Nuclear Fuels about the economics of the THORP reprocessing plant, elicited a series of replies to the effect that the departments had not themselves been supplied with the requested information. Providing this response however took the Department of the Environment 35 working days, the Cabinet Office and the Department of Trade and Industry 36 days each, and the Treasury 41 days.

* The Home Office failed either to reply to or acknowledge a request to it for internal guidance on an immigration matter, despite a reminder letter. Both the original request and the follow-up were sent to the Parliamentary Under Secretary of State. When the Campaign later contacted the minister’s private office by telephone, 41 days after the request had been received, it confirmed receipt of both letters, but said the request had been forwarded to the Foreign Office, which was responsible for the guidance in question. The Foreign Office told the Campaign it had no record of having received the request from the Home Office. A copy of the application was then resubmitted directly to the FCO which provided the requested information 20 working days later.

Delay in carrying out internal review

Before a complaint can be made to the Ombudsman the department must be given the opportunity to review its own decision. The code contains no target time for conducting such reviews. In practice they may involve substantial delay:

* A request for internal review by the Health and Safety Executive took 105 working days. The problem on this occasion may have been exacerbated by unusual factors, including the HSE’s move to new premises during the time in question. A subsequent internal review was completed by the HSE much more rapidly.

* A request to the Lord Chancellor’s Department was turned down after 27 working days. The internal review of this decision, which eventually confirmed the original refusal, took a further 55 working days (ie 11 calendar weeks).

* A request to the DTI (already mentioned above) was turned down after a delay of 36 working days. The first stage of the DTI’s subsequent internal review took 34 working days, and the second stage review an additional 63 working days. Excluding the time taken by the Campaign in preparing the appeals, the request and the appeals were under consideration by the department for a total of 133 working days, or six calendar months – an unacceptable delay.

Multiple-stage internal reviews

‘Internal review’ in most departments involves a single stage of appeal to a senior official. However, as the above example suggests, a few departments have introduced more elaborate procedures, involving repeated appeals. Both the Department of Trade and Industry and the Ministry of Defence have (or had) a two stage process of internal review, which requires the applicant first asking the official who took the original decision to review that decision. If the outcome is unsatisfactory, the applicant must make a second internal appeal to a more senior official. Since the Ombudsman will not normally entertain a complaint until a department’s internal appeals process has been exhausted, the effect of these procedures is to place still greater obstacles between the applicant and the Ombudsman.

The Campaign has made representations on the two-stage appeal process to both the DTI and MOD, and we understand that the DTI has now revised its procedures and adopted a single stage appeal. However, even a single stage review which extends over 63 working days – as in the above mentioned case – is excessive.

The most elaborate multi-stage appeals procedure has been introduced by the Northern Ireland Office, whose procedures may involve as many as four separate stages of internal appeal. The NIO’s Environment Department’s procedural note states:

“The formal procedure for internal review of appeals against disclosure of information is as follows. Initial appeal will be to the Grade 7 in charge of the Branch which dealt with the request. Heads of Division may wish to review the decision to deny access to the information at this stage in conjunction with the Central Management Branch, or as necessary with the Departmental Solicitor. If not satisfied with this review, the applicant will be entitled to ask the Under Secretary with lead policy responsibility to review further the decision to refuse access. If the applicant is still not satisfied the Permanent Secretary may be asked to arbitrate and in extremis the case may be referred to the Minister. If the applicant is still not satisfied the Ombudsman may be involved in the reviewing the Department’s decision.”

These procedures are no doubt intended to give applicants the maximum opportunity to challenge decisions. In practice they represent an endless series of internal reviews, each of which must be completed before the applicant may approach an MP with a complaint for forwarding to the Ombudsman. We doubt whether anyone would ever succeed in taking a complaint to the Ombudsman under such conditions.

Such excessive bureaucratic requirements tend to contradict the government’s assertion that the code is more informal and less bureaucratic than freedom of information legislation.

Nothing in the ‘Open Government’ white paper, the code itself or the Cabinet Office guidance on the code suggested that the government envisaged anything other than a single stage of internal appeal. We suggest that internal appeal should be limited to a single stage.

We also suggest that departments be required to complete the internal review within a fixed time, and that if they fail to do so the applicant should then be entitled to complain directly to the Ombudsman without further delay.

Internal review may be useful to departments, since it offers them the opportunity to correct a mistaken decision without the inconvenience of an external investigation. However, to the applicant internal review may be useful only if the department deals with it more speedily than the Ombudsman, and therefore offers the prospect of obtaining satisfaction more quickly than by complaining to the Ombudsman.

If departments take almost as long to deal with internal review as the Ombudsman would take to deal with a formal complaint then there is no value to the applicant in this procedure. The applicant’s interests would be better served by a right to complain directly to the Ombudsman in the first place.

The precedent for this system of internal review appears to be the Australian Freedom of Information (FOI) Act. However, internal review is not a feature of all countries’ FOI legislation. It is not required under the Canadian act where applicants can approach the Information Commissioner directly, nor in the New Zealand legislation where the Ombudsman can be approached directly. These countries’ FOI laws appear to provide a more user friendly right of access than the code of practice.

In Australia, under section 55(2) of the FOI Act, applicants can complain to the Administrative Appeals Tribunal, which enforces the legislation, but must first have applied for internal review. However, they are not obliged to wait indefinitely for the review’s completion. If the review has not been dealt with within 14 days, the applicant is then free to complain to the tribunal.

We believe a similar provision, with a fixed time period, should be adopted under the code. This would allow departments the opportunity to avoid an external investigation provided they acted reasonably quickly, while offering applicants the opportunity to go directly to the Ombudsman if the process of internal review had not been completed within a set period.

The “MP filter”

Given the potential obstacle of internal review, the fact that applicants cannot complain directly to the Ombudsman but must ask an MP to forward their complaints, adds a further complication to the complaints procedure.

The “MP filter” may not be a problem for organisations and individuals who are already in contact, and enjoy reasonable relations, with their own (or another) MP. It may well be an obstacle for others, who may find the prospect of approaching an MP intimidating and a further source of delay.

We appreciate that the committee has recently concluded that the MP filter should be retained, and helps to keep MPs in touch with problems affecting their constituencies. However, we wonder whether the balance of argument is the same for open government complaints as for maladministration. Only people who have personally suffered injustice can complain about maladministration; such a complaint almost by definition will relate to the personal circumstances of a constituent, which presumably may sometimes reflect the problems of other similarly placed constituents. However, open government requests, need not refer to the applicant’s personal circumstances at all. They may be prompted by wider concerns – the government’s approach to the European Union, for example – which have no particular local implications. We would prefer it to be left to individuals to decide whether or not to seek an MP’s help in obtaining the information, and allow them to approach the Ombudsman directly if they wish.


Most of the requests which the Campaign itself has made under the code have been dealt with free of charge. However, departments’ published fee schedules permit what we would regard as potentially prohibitive fees to be charged; and we are aware of examples of such charges being made.

In November 1994, the Campaign published a survey of departments’ charging policies. A copy of the survey with the accompanying press release is attached.

We reported that three departments (MAFF, Inland Revenue and the Foreign & Commonwealth Office) charge most applicants a basic fee of £15, which is non-returnable even if no information is supplied. Additional hourly charges may be made if the request is complex.

Such fees, charged even for simple requests, are not in our opinion compatible with a disclosure scheme whose stated aim is to open up the Whitehall culture. Far from doing so, these application fees suggest that requests for information are regarded as a costly inconvenience.

They may also be a practical deterrent to users. It is significant that 9 out of the 90 requests made to the Inland Revenue in 1994 were withdrawn after applicants were notified of the fees [OPSS, 1994 report, page 18, footnote 22].

Under the OPSS guidance, fees should not be charged for information which was previously available without charge or for information falling within a limited number of classes. Both MAFF and the Inland Revenue point this out in their leaflets. However the Foreign Office leaflet fails to do so, implying that every formal request under the code attracts a fee.

The Campaign itself was asked to pay £15 by the FCO for seeking information which, under the code, should have been provided free of charge. The fee request was withdrawn, when the Campaign pointed out the mistake. Other users, less familiar with the code, may be unlikely to challenge such charges. In our view the exceptions to the charging policy should be acknowledged in the Foreign Office’s own publicity.

Most departments allow some free time before charges are made. The amount varies from as little as one hour to as much as five hours, depending on the department.

Moreover, after the free time has expired some departments charge only for any subsequent staff time. Others charge for all the time spent, including the initial hours. The DTI and Welsh Office waive fees below £100, but charge the full costs – including the first £100 – if a request is costed at over £100.

This means that a request to the DTI involving three hours of senior staff’s time would be free: but if the request took four hours, the charge would be £136. (If less senior staff are involved, the first 7 hours would be free but the 8th hour would cost £112.)

Departments may feel that fees are necessary because they fear they may otherwise face potentially limitless demand for information. However, the danger at present is precisely the opposite: that the code will be barely be used at all. Moreover, the code itself provides multiple safeguards against unreasonably burdensome demands. Exemption 9 permits requests to be refused where they are:

“vexatious or manifestly unreasonable or are formulated in too general a manner, or which (because of the amount of information to be processed or the need to retrieve information from files not in current use) would require unreasonable diversion of resources.”

Using fees to control demand in practice favours commercial organisations, who generally can afford to pay. The Campaign’s press release on fees cited examples of charges of hundreds or even thousands of pounds. These would clearly put information beyond the reach of the wider public and most voluntary organisations.

[Note: the Campaign’s press release also referred to the fees of £50 and £100 a page charged in certain circumstances by the National Rivers Authority. We understand that the NRA has subsequently revised its charging schemes, so the fees referred to in the press release may no longer be current.]

We are aware of one request, from the Shetland-based Northern European Nuclear Information Group (NENIG) which asked the DTI a series of detailed questions about nuclear fuel reprocessing at Dounreay. The department asked for an advance payment of £325, which the group paid – a substantial investment for a small organisation. Despite repeated reminders, no information was supplied for 3 months. NENIG say they were told at various times that their request was with the department’s lawyers and, later, with the minister. When an answer was finally provided, it consisted of just two and a half pages, with none of the actual documentation the group had requested. NENIG maintain that two thirds of this was already public knowledge. They estimate that this information has cost them more than £5 a line. The department’s internal review upheld the limited disclosure and the charges, and suggested that in fact the costs to the department were much greater than those charges. NENIG were so dismayed at this response that they initially decided to abandon the matter altogether. However, they have since taken the issue to the Ombudsman.

There is also considerable discrepancy between departments’ charging policies. The most generous, relatively, appear to be the Lord Chancellor’s Department, the Public Records Office and parts of the Northern Ireland Office which allow five hours free work, and then charge only for any additional time; and the Cabinet Office, the Department of National Heritage and the Scottish Office, which waive the first £100 of all fees. At a minimum we would wish to see all departments bring their charges into line with these. However, we would also like to see fees being waived altogether where disclosures are in the public interest.

The OPSS guidance on the code does provide for fees to be waived, but only in limited circumstances, where disclosure is necessary for the “fair and accountable” performance of a department’s functions.

This potentially broad term is given an unreasonably narrow meaning by the guidance, which appears to limit it to Citizen’s Charter information such as details of benefits, service standards and appeal rights:

“Departments should not charge for the provision of information which it is necessary for the public to have as part of fair and accountable performance of their functions. Information explaining:

benefits, grants, rights and entitlements;

the standards, and availability of services;

the reasons for administrative decisions made in the applicant’s case;

the ways in which the citizen may exercise rights to appeal or complain about a decision;

regulatory requirements affecting affairs of a business, or common interests;

the main points of existing departmental policies or initiatives;

should usually be available free of charge.” [Guidance, Part 1, paragraph 71]

We believe that a much broader concept of accountability is called for in this context, and that fees should be waived where there is a public interest in the disclosure of the information.

A basis for this may derived from the code’s aims, one of which is:

“to improve policy-making and the democratic process” [paragraph 2].

Referring to this passage, the OPSS guidance adds:

“The public interest in disclosure is particularly strong where the information in question would assist public understanding of an issue subject to current national debate, or improve the transparency and accountability of a particular function of Government” [Part I, paragraph 3]

We think the code should provide for fees to be waived where such a public interest exists, while continuing to permit fees to be charged to commercial organisations or for requests not meeting this test (or not involving personal files). In broad terms this is what is done under the US Freedom of Information Act.


The code applies only to those bodies within the Parliamentary Ombudsman’s jurisdiction. While this includes virtually all government departments and many of the more important central government bodies, there are many important omissions.

The Ombudsman has recently referred to the “large number of public bodies” not subject to the code [HC 91, paragraph 13]. It is unlikely that anyone devising an open government scheme from first principles would exclude from its scope bodies such as the police and police authorities, the nationalised industries, the Atomic Energy Authority, the Monopolies and Mergers Commission, the Civil Aviation Authority, the Crown Prosecution Service, the Bank of England, the National Curriculum Council, the Broadcasting Standards Council. Training and Enterprise Councils and many others.

In addition, there are is a substantial list of matters, described in Schedule 3 to the Parliamentary Commissioner Act 1967, which the Ombudsman is prohibited from investigating. It is not clear whether these restrictions will also prevent the investigation of complaints about the withholding of related information. If this proves to be the case, a number of extremely important additional areas will be excluded from the code’s scope.

The proposed NHS code

The Parliamentary Ombudsman cannot investigate open government complaints relating to NHS bodies, such as trusts and health authorities, but such bodies are to be subject to a separate NHS code, supervised by the Health Service Ombudsman. A draft NHS code has been issued for consultation, and we have been particularly critical of its proposals. A copy of our response to the Department of Health is attached.

In our view, the draft NHS code is markedly inferior to the existing central government code. In particular:

* the central government code applies retrospectively, whereas the NHS proposals would only apply to information collected after the code comes into force

* most of the central government code exemptions contain tests of harm; these have been omitted from many of the corresponding draft NHS exemptions. Thus, while the central code exempts information whose disclosure would “prejudice… commercial or contractual activities” the NHS proposals omit the word “prejudice” and exempt all information about commercial or contractual activities, even if disclosure would cause no harm. This appears to exclude the basis of the internal NHS market reforms from the scope of the code. Other exemptions in the NHS proposals have been similarly broadened.

* the central government code contains a valuable provision, allowing the Ombudsman to rule that even exempt information should be disclosed where there is an overriding public interest in openness. This has been omitted from the NHS proposals.

Ironically, while trusts and health authorities would be subject to this inferior code, the Department of Health itself is subject to the central government code. This means that NHS bodies would be entitled to withhold information which, if held by the department, would have to be disclosed.

Sometimes, information on a single issue will be divided between NHS bodies and the department, with some information held centrally and other held locally. An applicant may apply to both for the information, but be treated under two separate and inconsistent codes.

Any complaints about non-disclosure will be divided between the Parliamentary Ombudsman and the Health Service Ombudsman – but both posts are currently held by Mr William Reid. As a result, he could find himself recommending disclosure of the information held by the Department of Health but unable to recommend disclosure of equivalent information held by the NHS body. This may lead to inconsistent decisions for which no intelligent rationale can be offered, placing the Ombudsman himself in a quite unreasonable position. We believe the answer is to bring the NHS code, which has not yet been issued in final form, into line with the provisions of the central government code.

Statutory secrecy provisions

Because the code has no statutory basis, it cannot override the many statutory restrictions on disclosure contained in legislation. The Open Government white paper identified some 400 of these. Many are unobjectionable and protect the privacy of individuals. Others prohibit the disclosure of all information obtained under particular powers, regardless of whether, for example, personal privacy or trade secrets are involved. One instance is section 118 of the Medicines Act 1968, which prevents the medicines licensing authority from revealing information about the safety of pharmaceuticals which it has received from manufacturers.

In 1992 the government began a review of these provisions, and committed itself to amending or repealing those that were excessive. Having completed the review, it appears to have concluded that the exercise was too complex, and the benefits too uncertain, to be worthwhile. Instead it has resolved that any future statutory restrictions must be more narrowly drafted. Existing statutory restrictions therefore remain in place, and access requests under the code will not be capable of securing information to which they apply.


Although so far only a small number of cases have been dealt with by the Ombudsman, these indicate that he is taking a tough line with departments over the code. The systematic way in which the Ombudsman’s interpretation of the code’s exemptions are being published is also welcome, and suggests that an important body of precedents will gradually become available.

However, the most obvious question about a non-statutory code is whether it will be complied with.

We accept that departments will come under considerable pressure to comply with the Ombudsman’s recommendations and will wish to avoid the public criticism which failure to do so may involve. However, we are not convinced that such pressures will always be sufficient.

Before the code’s introduction Mr Waldegrave cited the Ombudsman’s record of securing compliance over the last 25 years. At a seminar in October 1993 he asked:

“Where are the examples of a finding on maladministration or recommendations for compensation being ignored? The Ombudsman’s analysis has sometimes been disputed, but recommended remedies have not been held back”.

Yet only in the last few weeks, such examples have occurred. The Permanent Secretary at the Department of Social Security recently turned down the Ombudsman’s suggestion that compensation be paid to a man for the distress and anxiety caused to him and his wife after he had been wrongly accused of fathering another woman’s child by the Child Support Agency [HC 135, January 1995]. Last month, the Permanent Secretary at the Department of Transport rejected the Ombudsman’s findings of maladministration over its failure to compensate householders whose property had been blighted by the Channel Tunnel Rail Link [HC 193].

It is therefore entirely possible that the government may reject some of the Ombudsman’s disclosure recommendations.

We have argued that the open government code in fact provides greater potential for conflict between the Ombudsman and government than conventional maladministration cases. This is partly because applicants do not have to show that they have personally suffered injustice before making a complaint under the code. This opens the door to a much wider range of complainants than arise in conventional maladministration issues. Information may be sought by people actively participating in discussion of, or challenging the government over, policy matters. Complaints to the Ombudsman may be more likely to involve matters of current public interest or contention.

Moreover, the code contains a large number of so far untested concepts, for which there no equivalent to the body of maladministration “case law” now built up over nearly 30 years. There is considerable scope for disagreement over the interpretation of many exemptions, particularly where they refer to the potential harm to, say, defence, or to the candour of internal discussion. Finally, if a department is pressed for information on an issue on which it feels vulnerable, and where disclosure might highlight embarrassing flaws in a policy, it may resist disclosure in order to save face.

The government suggests that, should an Ombudsman recommendation be ignored, the remedy would be for him to make a report to Parliament which might then taken up by this committee. The prospect of an investigation by the committee may make the government more reluctant than it would otherwise be to reject a recommendation, and we recognise that the committee has warned departments to expect a grilling if they do reject a recommendation [HC 33, paragraph 88]. Departments will clearly anticipate this prospect before deciding not to accept the Ombudsman’s findings. But if they do decide to reject a recommendation, they will presumably have foreseen, and decided to resist, any additional pressure from the committee.

Clearly, it will be easier for the government to resist the recommendation of a select committee than to defy the order of a court or tribunal. This is of course the case for a statutory right of access. Under the Right to Know Bill, introduced in the Commons in 1993, enforcement would have been in the hands of an Information Commissioner whose rulings would be legally binding, subject to a right of appeal to an Information Tribunal. The arrangement was designed to keep disputes out of the courts, with all the costs that they involve, while retaining an enforceable obligation. Failure to comply with an order of the Commissioner or Tribunal would have been treated as contempt of court.

We do not deny the potential usefulness of the code. It is a step in the right direction, but it suffers from a number of substantial defects. We continue to believe that legislation covering the whole public sector, providing access to documents not information, and an accessible and enforceable right, remains necessary.

Scroll to Top