Australia’s ‘great surprise’ for the United States: Negotiating the 2002 Security of Information Agreement
This article originally appeared with this title in Freedom of Information Review 106 (August 2003), pages 50–51.
In June 2002 Australia’s Foreign Affairs Minister, Alexander Downer, announced that he had signed a new agreement on the exchange of classified information with the United States. In a public statement, Downer implied that the agreement was needed to account for “advances in information technology” since the signing of an earlier US-Australian pact in 1962.[1] However, internal documents released by the US Department of Defense show that the new agreement was prompted by very different concerns.
The US-Australia pact is one of over fifty bilateral security of information (SOI) agreements that have been negotiated by the US Department of Defense over the last half-century. Under an SOI agreement, governments agree to take a variety of measures to control the handling of sensitive information received from the United States. The US practice of negotiating SOI agreements was formalized in National Security Decision Memorandum 119 (NSDM 119), issued by the Nixon administration in June 1971, which prohibits the sharing of classified military information with a foreign government that has not signed a legally binding SOI agreement.
These SOI agreements can have important implications for citizens. For example, security clearance procedures that are needed to comply with the requirements of an SOI agreement could affect the ability of Australian citizens to hold government employment. (In the 1950s and 1960s, NATO SOI rules effectively barred homosexuals from key government positions.)
SOI agreements can also collide with national right-to-information policies. The US-Australia agreement absolutely bars the disclosure of shared classified information to non-governmental actors without the consent of the originating government — a flat prohibition on disclosure that clashes with the harm-based approach to disclosure upon which right-to-information laws are often based. The agreement also states that disagreements about the disclosure of shared information “shall not be referred to a national court, to an international tribunal, or to any other person or entity for settlement.”[2] This could conflict with appeal provisions in Part VI of Australia’s Freedom of Information Act.
Few Australians would have known about the 1962 SOI agreement, which was itself a classified document. This was standard practice at the time. The US-UK SOI agreement, signed in April 1961, was only declassified and publicly acknowledged in March 2001. Canada denied the existence of its agreement with the United States until it was declassified in response to a request from the US government in November 2002.
In Australia, the classified status of its SOI agreement with the United States created distinctive problems. In August 1999 — almost forty years after its signing — the Australian Defence Department told the US Department of Defense (US DOD) that it did not view the 1962 agreement as legally binding under international law. According to documents released by the US DOD, the Australian government said that it had never undertaken the internal procedures necessary to make the agreement legally binding, and regarded the agreements as “morally and politically binding only.”[3] Australian officials told the US DOD that the classified status of the 1962 pact had complicated the process of making it legally binding — perhaps because some officials lacked the clearance needed to approve the document.
This caused serious concern within the US DOD. In November 1999, a US DOD official told US Deputy Secretary of Defense John Hamre that the Australian message “came as a great surprise to us, since all U.S. international agreements concerning security of information have always been done on a legally binding basis. It is also of great concern to us, since these agreements establish the fundamental commitment to protect classified information of the other government. DoD shares a great deal of highly classified information with Australia.”
Australia was the only country whose SOI agreement with the US was not legally binding. “It is imperative,” the official said, that a legally binding agreement be established as soon as possible.
Deputy Secretary Hamre was advised to use an upcoming visit with Australian officials to “urge [them] to negotiate and conclude the replacement agreement quickly.” In the meantime, US DOD staff planned to evaluate “disclosures of classified information to Australia program by program” to ensure that stopgap measures were in place to provide legally binding assurances before disclosures were made.
The US DOD also had other bureaucratic difficulties. The 1962 agreement covered the exchange of all classified information, but NSDM 119 only required a legally binding agreement for the exchange of classified militaryinformation. The State Department has insisted since the mid-1980s that new SOI agreements be limited to the military sphere alone, and in November 1999 it was expected to resist the US DOD’s request to negotiate a new agreement equal in scope to the 1962 agreement. In the end, however, the State Department did not block US DOD’s plan. As a result, the new US-Australia agreement is actually broader than required by NSDM 119.
However, there may have been other complications in talks about a replacement agreement. A July 2000 memorandum from US DOD said that formal negotiations for the new agreement would “take place in Canberra during the week of 17 July 2000.”[4] This proved to be a highly optimistic forecast. In fact, negotiations went on for nearly two years until the new agreement was signed in June 2002. US DOD memoranda do not provide an explanation for the protracted discussions.
That the new SOI agreement is publicly accessible — unlike its 1962 predecessor — is a small step forward. Nonetheless, greater transparency about the negotiation of such agreements is clearly required. As defence and intelligence services collaborate more closely, these agreements will grow in number, and the constraints which they impose on national right-to-information policies may become tighter. This is an area that has been wholly neglected by academics and activists interested in governmental openness. One first step might be an inquiry through Australia’s Freedom of Information Act about the Australian view of negotiations over the 2002 SOI agreement. There are other opportunities for study as well. Since 1996, Australia has negotiated at least four other SOI agreements — with South Africa, Canada, Singapore and Denmark.
[1]Australian Department of Foreign Affairs and Trade. 2002. Media Release: Australia-US Sign Pact on Classified Information. Canberra: Australian Department of Foreign Affairs and Trade. June 25, 2002. http://www.foreignminister.gov.au/releases/2002/fa093_02.html. Accessed September 2, 2003.
[2]Article 18 of the 2002 United States-Australia Security of Information Agreement. The text of the agreement can be read at http://www.austlii.edu.au/au/other/dfat/treaties/2002/25.html. Accessed September 2, 2003.
[3]This and following quotations are drawn from the following source: Memorandum from Director, International Security Program,s to the Deputy Secretary of Defense, on the Australia General Security of Information Agreement, November 16, 1999. Memorandum I-99/014889-IS. The memorandum and other documents released in response to US DOD FOIA request03-F-0108 (April 8, 2003) can be downloaded from the author’s website, http://www.foi.net.
[4]Information Paper on U.S./Australia General Security of Military Information Agreement prepared by Mr. P.R. Batten, US DOD International Security Programs, July 6, 2000.
