Remarks on security classification appeals

Thursday, April 1, 1999

I am delighted to have this opportunity to discuss the work of the ISCAP at this conference marking National FOIA Day.

After two years of review, President Clinton decided in 1995 that it was time to revisit some of the rigid classification categories that have defined our policy for over 50 years. The resulting order marked a significant change in the assumptions about what should remain classified on national security grounds. Before ISCAP, there had been no interagency body to review an agency’s decision to keep information classified since 1978.

I think it’s fair to say that the first members of the ISCAP — drawn from Justice, State, Defense, CIA, NSC, and Archives — were skeptical that they could move this huge glacier — this battleship of secrecy, as the Attorney General sometimes refers to it — in the direction of maximum responsible disclosure. The skepticism was deeply rooted in two realities: first, the culture of secrecy that has governed our classification policy for over 50 years. And, second, the reality that threats to our national security have grown infinitely more complex and fragmented since the end of the Cold War. Skepticism reigned even in the face of a quite dramatic reversal of the resource burden for keeping information over 25 years old classified. Under prior orders, agencies had to expend resources in order to declassify older information. Inertia favored continued classification. Now, agencies have to justify retaining classification, by demonstrating how older, historical information falls within one of the narrow exceptions to automatic declassification after 25 years.

Would the same agencies that had so fiercely protected their secrets for so long conform to the new Order? Would members of ISCAP with line or supervisory responsibility within the very agencies whose decisions were being questioned be able to wear a second, “independent” hat? Or would they find ways to bend the new Order to fit their existing practices? Believers in the Great Myth — that government cannot reform itself — were not optimistic.

I think ISCAP has slain that myth. To date, all ISCAP decisions have involved appeals of agency decisions to retain classification in response to mandatory review requests permitting members of the public to seek declassification of specific information without making a FOIA request. A vote of a majority of those present is necessary to overturn an agency’s classification decision.

Since our first meeting, in May 1996, through January 1999, ISCAP has voted on a total of 163 appeals. Of these, we have fully declassified 83 documents, or 51 percent. We have declassified significant new information in an additional 52 documents, or 32 percent, while entirely upholding agency decisions to retain classification of 28 documents, or 17 percent. Thus, we have declassified new information in 83 percent of the documents we have voted on — information that was kept classified at the highest levels of appeal within the agencies. Nearly all of these cases involved information from the State Department, CIA, or the Pentagon. The issues presented include foreign relations, intelligence sources and methods, and military plans and weapons systems, in roughly that order of prevalence.

There’s more bad news for the Great Myth: nearly 85 percent of our votes have been unanimous. This means that ISCAP members from the State Department, the Defense Department, and the intelligence community often vote to declassify information that their own agencies had kept classified. ISCAP members are in fact scrutinizing their own agencies’ classification decisions in an independent and unbiased way.

I will mention briefly some the subject matters that have been the subject of ISCAP declassification decisions.

  • Documents from the Eisenhower, Kennedy, and Johnson Administrations on the deployment and potential use of nuclear weapons in Europe. Among the subjects discussed were the targeting of weapons against the nations of the Warsaw Pact; command and control of nuclear weapons in emergency circumstances; and the relative authorities of the chairman of the Joints Chiefs of Staff vis-á-vis the Supreme Allied Commander in Europe. With the exception of a small portion or two, the ISCAP voted to declassify this information.
  • State Department messages from June and July 1967 (the aftermath of the “Six Day War”) discussing the West Bank and assessing Israeli capabilities and intentions regarding nuclear weapons.
  • Documents, dating from 1974-76, on South Korean nuclear processing and the potential development of nuclear weapons and missiles by that nation. In these two cases also, we declassified substantial new information, while keeping some parts of the documents classified to protect current foreign relations or intelligence-gathering efforts.

But ISCAP’s caseload represents only a small fraction of the overall declassification program. ISCAP may be a successful laboratory for reinvention of classification policy, but how can its work affect day-to-day, government-wide declassification review?

This is a fair question and a big challenge. Certainly we can continue to issue our periodic communiques and make available our decision database on computer diskettes. Steve Garfinkel and I periodically speak to interested audiences, including agency historical advisory committees, about ISCAP’s work. Our most recent communique is available on the Department of Justice web site:

But is there more? We are now looking at two additional means of leveraging ISCAP’s decisions.

1. The first means is through our review of proposed agency classification guides. This is a major ISCAP responsibility under the Executive Order, and it is still ahead of us. Agency heads who wish to exempt specific information from automatic declassification after 25 years must submit their proposed exemptions for ISCAP review and approval. We expect most submissions to take the form of proposed classification guides, identifying what types of information could be kept classified and, in most cases, specifying a date or event for declassification. The Order’s implementing directives require these guides to “state precisely the elements of information to be exempted from declassification,” [32 CFR § 2001.53(b)(6), thereby avoiding broad generalizations.

Agency heads are required under the Executive Order to make these submissions by October 1999. To avoid a last-minute stampede, we are strongly encouraging agency heads to make their requests as soon as possible. The ISCAP staff, with working-level assistance from member agencies, has begun to review submissions to date. Agencies will get feedback on this review before the ISCAP takes any formal action.

It won’t be easy for the agencies or for ISCAP to produce a series of workable classification guides. But the upside potential is considerable. We have the opportunity to state what kinds of information at each agency require protection and how long such information should stay classified. This will allow ISCAP to affect the declassification of many thousands of documents beyond what it sees directly on appeal, and enable agencies to make more consistent and efficient determinations.

2. The second way for ISCAP to leverage its work is through outreach to the broader declassification community. Earlier this month, through the leadership of ISCAP’s Executive Secretary, Steve Garfinkel, Director of the Information Security Oversight Office, we tested a prototype of a more ambitious outreach. We had groups of declassification reviewers work through selected case-studies drawn from ISCAP appeals, and then compare results and rationales with those of other groups and with the ISCAP’s decision-making. The participants in this prototype — about 45 in all — were drawn from the declassification staff of the National Archives; the interagency Classification Management Committee, chaired by ISOO; and the intelligence community’s multi-agency External Referral Working Group, which works toward consistent and efficient review of intelligence community equities. Six ISCAP documents, representing military, diplomatic, and intelligence-related issues, were discussed.

I understand the general reaction was quite positive. We will closely examine the evaluations from our test audience and assess whether it makes sense to continue. If it does, we will try to maximize the effectiveness of future presentations — for example, by selecting case-studies with a view toward issues that recur at a specific agency.

Now for two of the thorniest issues that present recurring questions for ISCAP. One of the recurring categories is “foreign government information” or “FGI.” Governments routinely share a great deal of information — through diplomatic, military, intelligence, law enforcement, and other channels. Very often, there is an understanding — express or implied — that the receiving government will not turn around and disclose the information to the public.

Under the prior Executive Order (12356), disclosure of FGI was presumed to damage national security. In practice, this meant that FGI was classifiable in perpetuity.

Executive Order 12958 removed FGI’s iconic status. Like predecessor orders, it identifies FGI as a separate category for original classification. However, it is no longer a “classifiable-forever” category. Now, FGI is declassified after 25 years unless it satisfies the same stringent criteria that apply to information of U.S. — origin. For example, FGI can stay classified for longer than 25 years if release would “seriously and demonstrably impair either relations between the United States and a foreign government” or “on-going diplomatic activities.”

Some old FGI will meet the new, stricter standards; some won’t. Applying the standards to particular cases calls for discretion: How old is the information? Through what channels was it received? Might declassification and release deter the foreign government from sharing valuable information today? Is the subject diplomatically sensitive, even now? Is the best solution to declassify as much as possible, while protecting the foreign government as source? Should the originating government be consulted? These have proven to be difficult questions. They are no formulaic answers.

In considering FGI, the ISCAP has benefitted greatly from the expertise and example of the Department of State. I understand that the State Department will sometimes declassify FGI based on the age and sensitivity of the information and our current relations with the foreign government. In other cases, the same factors will counsel classification for longer than 25 years. State sometimes will consult with a foreign government, after thinking carefully about whether and how to do so. Even if the consulted government says, “Please don’t declassify or disclose,” classification after 25 years is not automatic.

The second recurring category is, of course, “intelligence sources and methods.” Under the prior Executive Order (12356), disclosure of intelligence sources and methods was presumed to damage national security. Thus, like FGI, this category of information was classifiable in perpetuity.

Under the current Order, “information about the application of an intelligence source or method” may remain classified for longer than 25 years, provided that the criterion for damage to national security is also met. In other words, information in this category may warrant classification after 25 years, but there is no presumption it should stay classified.

Here again, the reexamination of a classification category requires painstaking review. Secrecy is integral to what we ask intelligence agencies to do. No other field has a greater need for confidentiality, to protect both U.S. personnel and foreign individuals and institutions. Disclosing a genuinely secret technique can render it useless.

On the other hand, even intelligence sources and methods can become less sensitive with changing circumstances and the passage of time. Whole chapters of recent history cannot be credibly written without reference to intelligence activities. As the Report of Senator Moynihan’s Commission on Protection and Reducing Government Secrecy observed, excessive secrecy can fossilize processes it is supposed to preserve. It can “undermine well-informed judgment by limiting the opportunity for input, review, and criticism” and the “type of scrutiny that might challenge long-accepted beliefs and ways of thinking.” (Report at pp. 7-8) And the intangible costs of government secrecy — public suspicion, distrust, and cynicism – also are considerable.

I have pointed to the large percentage of unanimous votes we have had on ISCAP. The split decisions, which frequently involved intelligence community equities, are also illuminating. To me, this suggests that certain intelligence-related issues are genuinely close calls. Do the names of human intelligence sources working for the Allies during World War II require continued protection? When, if ever, should a cooperative intelligence relationship with a foreign government be declassified? How can it be that the use of an intelligence method many years ago can still merit secrecy today? We’ve had more than one closely divided vote on that issue. Reasonable, conscientious people can reach different conclusions about these issues, and each appeal will have to be considered on an individual basis.

The challenges we face in the post-Cold War world are more complex than those we faced during the Cold War, whether from terrorist threats, the proliferation of weapons of mass destruction, or the on-going need to engage diplomatic channels in effective ways. With the stakes so high, we cannot afford to be reflexive in either direction when it comes to classification policy. I look forward to your continued scrutiny and engagement as the ISCAP experiment continues.