Protecting Information: A Practical View by Graham Dominy:

It is right, proper and healthy for matters of human and constitutional rights to be vigorously debated in a democracy and for issues to be robustly argued and defended by proponents and antagonists, but I cannot escape the conclusion that the debate on the Protection of Information Bill would be better served if it were grounded more in reality than in hyperbole. I propose to discuss three issues in this article: Why a Protection of Information Bill is needed; how security and archival information classification systems work and how an Information Access regime can be better monitored and adjudicated.

As National Archivist it was my duty to serve on the Classification and Declassification Review Committee, the purpose of which was to investigate the problem of the masses of documents classified as secret and which have survived the destruction of apartheid secrets ordered in 1993 – 1994. The Review Committee conducted a wide ranging investigation which included interrogating the views of security specialists from the old order, from the liberation movements, from the new transformed security and intelligence agencies, as well as visiting the Stasi Archives in the Former East Germany, Hungary, the Czech Republic, the secret service in Canada and the CIA in the USA.

Ronnie Kasrils, former Minister of Intelligence Services, has cogently outlined the origins of the bill and given a neat contextualisation of the issues involved. This is, perhaps, the closest we can get to the public views of influential sections of the intelligence community. My purpose here is not to repeat or comment on what he has said, but to give you a practitioner’s or an archivist’s view.

The Bill of Rights in the Constitution states that “Everyone has the right of access” to state information and information held in the private sector “required (my italics) for the exercise or protection of any rights”: it should be noted that that this is an individual not a group right. The constitution then charges Parliament to pass legislation to give effect to this right and concomitant matters. In other words the right of Access to Information is not an absolute right; it is what is best called a mediated or conditional right.

The Promotion of Access to Information Act (PAIA) was promulgated in 2000 to give effect to this right. It does not mention the classification of information at all, nor does it mention the National Archives Act at all. The reasons for this are two-fold: as PAIA gives effect to a major constitutional right it “trumps” other legislation and other legislation needs to be aligned with it; also its prime purpose is to govern the Information Access regime and set out the framework within which this regime operates. This includes outlining the circumstances under which some information must, or may, be withheld.

It does not speak at all to the management of the type of information which must or may be withheld.  It is, therefore, very difficult to understand the reasoning which asserts that the PAIA exemptions mean that we do not need the Protection of Information Bill.

The Protection of Information Bill is intended to guide government departments and “organs of state” on how to manage information that needs to be protected and sanction those who flout the protections. Currently we are operating under the apartheid Protection of Information Act and its regulations which are largely unconstitutional. Much of the furore around the current Bill relates to the alleged peril that lurks if information has to be declassified before it can be released. Hard as this might be to believe, it happens in almost every democratic country I can think of off hand (from Australia and Canada via Germany to the UK and the USA). It is a time consuming, resource-intensive and laborious process and, believe it or not, the Bill proposes an agency that will alleviate this burden, the National Declassification Database which will be situated at the National Archives.

The Protection of Information Bill generally articulates and cross-references fairly well with PAIA and with the National Archives Act. We in the National Archives and Records Service regard records classification systems as our daily bread and butter. Although our legislation pre-dates PAIA, we have long argued that one way for each government agency to comply with PAIA is to publish their National Archives-approved filing plan in their PAIA manuals. Only the Intelligence agencies have been exempted from the requirement to publish PAIA manuals. Our Archives-approved records management systems also allow for the security classification of information and accommodate this within our guidelines.

One of the most controversial issues in the bill is who is authorised to classify information? There are many critics who believe that this should only be done by a Minister or very senior official with a rapid appeal or review instrument close to hand. While this might be the apogee of democratic aspirations, it is entirely impracticable and will surely cause the bureaucratic delays and backlogs it seeks to avoid.

What is classified information? It is a by-product of security, high policy or intelligence work. Let us take a hypothetical example: An intelligence officer hears from a terrified informant that there is likely to be an outbreak of xenophobic attacks in a certain area. If the state is to respond rapidly to protect lives and property; this intelligence officer cannot wait to get an appointment with a minister, general or director to classify the information as secret and disseminate it with the necessary speed and discretion through the decision-making structures. It needs a really rapid operational response which means that the secret information needs to be circulated within hours between local intelligence staff, their police colleagues, the army and emergency service workers.

Once the outbreak happens, or is forestalled, a lot of what is secret would probably come out into the public domain anyway. What would not come out and should never be allowed come out is the name and identity of the informant.

It is also costly and laborious to manage classified records. It requires special procedures; staff who are security-cleared and specific supplies, equipment and tracking mechanisms (a records equivalent of Netstar/Tracker). The Intelligence Agencies are geared up for it, but in ordinary departments the mechanisms do not work as smoothly (it is not the core business of such departments). Therefore every effort should be made to avoid unnecessarily classifying information and thus clogging up a creaking records management system.

I once had a PAIA request for the National Archives own secret files. It meant that I had to spend days going through piles of files in a locked safe. In the pile was an ominous sealed envelope; red seals “Streng Geheim” etc. I nervously opened the envelope to discover a letter from a former Director of Archives to the Minister of the time (Piet Koornhof) in which the Director mildly criticised the Public Works Department. Piet Koornhof thought that this should be kept a secret!

This might seem an extreme example but it serves to illustrate what can happen if decisions about the classification of documents are required to be taken by high ranking people not necessarily involved with the work on the ground on a day-to-day basis.

A classification adjudication system should only come into play when the state’s need for secrecy and the public’s right to know need are at odds. Currently one of the major weaknesses of our Information Access regime is the poorly articulated appeal mechanism. If somebody objects to my ruling (as National Archivist) that they cannot see a file, they must appeal to my Minister who then consults me and often takes my advice. The aggrieved applicant’s only recourse is then to the courts. This is expensive, time-consuming and does not necessarily guarantee a well-considered outcome because there very few specialists in this field.

Several countries, including Canada, have Information and Privacy Commissioners who are generally able to resolve over 80% of disputes at low cost. The remaining 20% of problem cases still need judicial adjudication, but this is a fair, cost effective and relatively speedy system of dispute resolution. In South Africa, the Human Rights Commission has a generalised oversight role, but is unable to intervene in the adjudication of individual cases.

The debate over the Protection of Information Bill is fascinating and crucial, but, inevitably, it has become an emotionally clouded slanging match. I hope that this outline of some of the practical issues involved will contribute to a more realistic level of discourse.

Dr Graham Dominy is the National Archivist of South Africa and he and several of his colleagues are currently studying online for the Information Access and Privacy Protection Certificate offered by the University of Alberta in Canada in preparation for the establishment of the National Declassification Database. This article is written in his personal capacity and does not necessarily represent the views of the Minister of Arts and Culture or the South African Government.

This article was first published in the Cape Times of 26 August 2010

Leave a comment