FAS Homepage | Government Secrecy | Gov Docs ||| Index | Search | Join FAS


By John W. Carlin
Archivist of the United States


Voice of America Auditorium
United States Information Agency
Washington, D.C.

December 5, 1996

I'm very pleased to be here today. And I want to thank the United States Information Agency for hosting this conference. Declassffication is a critically important topic in the world today, and we need forums like this one in which we can share our progress and our ideas with each other.

Just a took at recent headlines in newspapers around the country reinforces how important access to long-classified information can be. Classified files have led Holocaust victims to take on the Swiss banking establishment, have led historians to re-evaluate who knew what when in the early years of World War II, and have shed further light on President Kennedy's actions during the Cuban Missile Crisis. Yet far too many documents remain classified. And documents that remain classified beyond their time are a great burden on our democracy.

At the National Archives and Records Administration our mission is to provide ready access to essential evidence--evidence that documents the rights of American citizens, the actions of federal officials, and the national experience. The mission of the U. S. Information Agency is not so different; it's just aimed at a different audience. Like NARA, USIA provides ready access to the essential evidence emerging democracies and closed societies need to build civil and free societies. Without open and accessible records, neither one of us can fulfill our missions. And long-withheld and improperly classified records pose a threat to our missions and our government.

Our democracy was founded on the premise that an informed public is necessary for the proper functioning of government. Without access to information about what their government is doing, citizens are unable to make appropriate decisions about the people who represent them and the directions that government should take. This lack of information can lead citizens to distrust their government. People assume that if the government is not forthcoming, it's because it has something to hide--incompetency, corruption, scandal or something worse. And a democracy cannot function without a basic level of trust in that government on the part of those governed.

Our past failure to deal with the growing backlog of security-classified information has contributed to the present climate of hostility toward government. The steps we take to implement Executive Order 12958 can help restore the trust that was once taken for granted.

Now, let me say at this point, that although as Archivist of the United States I will stand for open and accessible records, I do not do so indiscriminately. I know that there is a place in our government for classification of highly-sensitive national security documents. And I take my responsibilities to protect classified information seriously. But we also can't use outdated and outmoded classification policies as excuses for keeping information from the American public and damaging the public's trust in government. We must be responsible--and proactive--when dealing with national security information.

For too long, declassification has been a low priority, except for agency historians. It's time-consuming, expensive, and seemingly doesn't contribute to an aaency's mission. As a result, we've accumulated huge backlogs of classified infomation. These backlogs generate Freedom of Information Act requests from those who want access to the information. These requests then divert resources from what little systematic declassification is being done and contribute to the further growth of the backlog.

Now, in the year since Executive Order 12958 took effect, I've watched with great interest, and some concern, the ways in wlich we in the Federal government are implementing the declassification provisions of the order. And I think we've all realized that the problem of declassification of older records is not a problem that can be handled by just one agency--the National Archives and Records Administration--or by each agency workin independently. It's a problem that can only be solved by all of us working together, coming up with ways to make our efforts complement each other, developing standard procedures, and searching for new methods or automated tools to make our work more efficient and effective. Most important is frequent interagency communication--through forums like this and through other, less fomal avenues--to share the procedures and tools that work best.

Actually, for the most part, the results of the first year have been good. Most agencies have put declassification programs in place and are making honest efforts to meet the goals for dcclassification specified in the Executive Order. You're consulting with the user community for advice on declassification priorities. And you're cooperating with us on the review of records in our custody.

Yet several difficult problems remain to be resolved, including the interagency referral process. But if we continue to share our progress and ideas, we will begin resolving these outstanding problems and develop a more effective declassification process overall.

I'm especially impressed with the level of interagency cooperation I've seen in the External Referral Working Group. This group has been very effective in not only providing a forum for discussion of mutual problems, but in actually getting a cooperative effort underway to start declassification review of high level materials in Presidential libraries. Their Remote Archives Capture Project is an excellent example of how, working together, we can create a more effective, efficient approach to the problcm of declassification at remote locations.

Let me explain. The materials at Presidential libraries present a unique declassffication problem because they are highly sensitive and usually contain equities of more than one agency, requiring multiple reviews. Through the remote capture project, the documents are scanned and delivered to an expert team of reviewers here in Washington, instead of each agency having to send separate teams of reviewers to each library. For the libraries it means quicker declassification--the libraries don't have to wait for different teams to make multiple reviews. And it means less handling of the documents. And it also may lead to declassification of more information. Documents that would have been withheld by a generalist reviewer might be released by a reviewer with expert program knowledge who would not be available for on site review at the libraries.

The results of the remote capture pilot projects at the Johnson and Kennedy libraries are promising. This appears to be a workable approach to declassifying documents in the Presidential libraries. And I greatly appreciate the efforts of all the agencies that have participated in this project thus far and urge any of you who may be interested to take part. The project will only succeed if agencies continue to support it with personnel or other resources.

Automation may hold additional promise for solutions to our declassification problems. We cannot continue to declassify in the old-fashioned way--vath a page-by-page review-- and hope to eliminate the backlog. We'll never have sufficient personnel for that. The Automation Working Group is looking at ways to make the process more effective government-wide. I'm encouraged by some individual agencies that are offering to share their automated approaches and to cooperate in developing new applications. And I expect the work of the Automation Working Group to open the door for further collaborations in this area.

Happily, collaboration has been a highlight of this last year. Your help has been essential to us at NARA because we do not have the resources to review our entire classified backlog without your assistance. And in some cases we still don't have the declassification guidance we need to review our records. We've made it our priority to identify those records that can be easily and quickly declassified so that agency personnel can concentrate on those records that most need their greater expertise.

And our results in this first ycar have been excellent. NARA staff examined 131 million pages, released 111 million, and set aside 20 million for further review. Agency personnel working at NARA reviewed another 10 million pages. Air Force, Navy, State Department, Department of Energy, Agency for International Development, and Arms Control and Disarmament Agency all have reviewers working at NARA on a regular basis. This level of effort will have to be sustained and probably increased over the next several years if we're going to meet the deadlines imposed by the order.

However, with all the success we can point to so far, there still remains at least one thorny problem--interagency referral of other agency equities. No agency's files are completely free of information received from other agencies. That means every agency wfll be trying to get the attention of other agencies to review information in their files. So far each agency has its own plan for making and receiving referrals. Some will simply notify other agencies of the existence of their equities, expecting the other agencies to come to them. Others plan to send copies of the documents, either in paper or electronic form. Some agencies will do a pass-fail review; others will redact. Some will review only for classified information. Others will do a full FOIA review. And everyone wants to deal with the records in their own possession before worrying about their equities in someone else's records.

The issue here is how can we minimize the number of documents that must be passed from agency to agency so no one is overwhelmed. This one problem has the potential to cause the declassification process to break down completely if we cannot come up with a solution. And yet agencies continue to insist on reviewing every document containing their equities.

To combat this problem we all must recognize that not every document containing another agency equity requires referral. Much of the information in our 25-year backlog is routine information, easily recognized as no longer sensitive. It refers to events long past, obsolete technology, or routine operations. We should be able to establish a baseline for types of information which do not need to be referred to the originator, but can be acted on by any trained declassifier.

One way to do that is by sharing guidance and delegating authority. Each agency would retain authority over those areas of particular concern to its mission. Everything in those reviewed areas still would need to be referred, but the vast majority of information could be acted on by the initial reviewer. Our experience at NARA over the last 20 years suggests that the amount of referrals could be reduced by 80 percent or more. NARA typically declassifies more than 90 percent of the pages reviewed using the guidelines provided by agencies.

I know some of you object to using guidelines on other agency information. It seems like too much extra work. But how much extra work will be involved in referring this information or handling other agency referrals? Some agencies object to having other agency personnel act on their information because they don't know what kind of training they've had. This can be overcome through the kinds of cooperative efforts at equities recognition training being sponsored by the External Referral Working Group.

Even if we manage to minimize interagency referrals, the amount of records that are subject to the automatic declassification provisions of the Executive order are beyond the resources of most agencies to review in the time allotted. Therefore, a risk management approach is the only practical approach. Balancc the cost of page-by-page review against the risk of inadvertent disclosure and focus on the records with the highest potential for problems.

Every agency needs to make a careful assessment of what 25-year-old issues, technology, and other secrets are most sensitive and determine where that information is most likely to be found. We need to concentrate on those records and accept that there may not be time to examine the rest as thoroughly. Security people and declassifiers need to work closely with records managers to prioritize these efforts.

Agency records managers often can provide information not only about the location of records, but also about the types of information in the records. The more information a declassifier has about the records, the better the decision he or she can make about how thorough a review is required. By using risk management techniques, we can minimize the possibility of inadvertent disclosure while making the most efficient use of our personnel and resources for declassification. Whatever changes are ultimately made in the Executive Order, we should never return to a declassification system that requires a total page-by- page review. Therefore, we must start now to incorporate records management into declassification risk assessment programs.

At NARA, a renewed emphasis on the management of records throughout their life cycle is a key component of our recently-issued strategic plan. I'm committed to putting more resources and attention on the front end of the life cycle, where record systems are designed, records are created, and filing systems are organized. I want to enlist federal agencies as partners with whom we share responsibility for managing records to assure the public of access to essential evidence. We plan not only to clarify the kinds of records that the law requires you to create and maintain, and to actively guide records management, but also to contribute to the design of recordkeeping systems. If we assist agencies with the management of their records from the beginning, we can save money for the government as a whole through more efficient recordkeeping and more responsive services.

And one way in which improved records management can assist the declassification process is by making it possible for security and declassification personnel to have accurate information about the volume of security-classified records an agency has, where the records are located, and what they contain. In many agencies this information was lacking when we started to implement the Executive Order, making it difficult to develop a realistic declassification plan.

We also must work together to establish clear and consistent guidelines for the classification and declassification of records throughout the government. Through continued collaborative efforts between NARA and other Federal agencies, we can make the sometimes mundane work of records management and declassification a force for real change in the way we do our work.

Now, let me raise a final caution. When talking about declassification, most of us still think in terms of paper. But classified records have been created in many formats-- audio tape recordings, motion picture films, microfilm, and electronic media. All of these present much greater difficulties for declassifiers than working with paper because of the requirement for special equipment to access the information. Scanning and on-line redaction systems might be adapted for review of microfilm. But at present the only way to declassify audio tapes, motion picture films, or videotapes is to listen to or view them. While these records are much smaller in volume than the mountain of paper rerords awaiting our attention, they are proportionally much more time consuming to review. We must concem ourselves now with developing effective methods and automated tools for declassifying records in these other formats.

Electronic records pose a special problem. We have growing quantities of computer generated records, ranging from millions of e-mail messages to vast scientific databases, all of which require new methods for appraisal, preservation, and public access. This includes the more than 5,000 magnetic tapes and hard disks made by the Executive Office of the President and the National Security Council under Presidents Bush and Reagan. Electronic databases cannot simply be "printed out," but must be electronically maintained to preserve their usefulness. Electronic maintenance means preventing inadvertent deletion of such data, transferring it from deteriorating tapes and disks, and making it accessible without keeping all the obsolescent hardware and software on which it was originally created.

When dealing with security-classified databases, it also means developing new methods for declassification, from scanning and on-line redaction to electronic referrals. At the same time, all of the original information must be preserved. And while relatively few electronic records fall within the 25-year bacldog, the decisions we make now about automated tools for declassification must be made with electronic records in mind.

In closing, records in all fomiats undergird not only the legitimacy of government, but also its accountability. However, for records to perform their proper function they must be accessible to the American people. At NARA, we take seriously our role as a public trust. We enable people to inspect for themselves the record of what government has done. We enable officials and agencies to review their actions, and help citizens hold them accountable for those actions.

But the National Archives of the United States is more than the records stored across the Mall at Archives I or in NARA's other facilities. The records you and other Federal agencies hold also are part of the National Archives, and we share the responsibility for ensuring that these records are held in trust for the American people.

Today, security-classified records make up a significant part of the holdings of our National Archives. Ensuring access to these records while protecting still sensitive information is one of the serious challenges we face in implementing the Executive Order. But because of this order we've made important strides for openness and access to government records. Let's continue to use this opportunity to attack the issues and problems that stand in the way of providing access. Let's expand the work that already has been started to reach common solutions to the challenges we all face. If we continue to collaborate, I believe there will be a tomorrow in which only that information which is truly sensitive will remain classified and in which the public's confidence and trust in government will be on the rise.

Thank you very much.

FAS Homepage | Government Secrecy | Gov Docs ||| Index | Search | Join FAS