The White House
Washington, DC 20500
Dear Mr. President:
I am pleased to submit the information Security Oversight Office's 1993 Report to the President.
The submission of this Report occurs while we are in the final stages of shaping the first post-Cold War security classification system: a system intended to reduce significantly the amount of information that we classify in the first place, and to reduce dramatically through declassification the amount of older classified information that has built up over the decades. The input for creating this revised system largely took place during the time period covered by this Report. During this phase of the process, many hundreds of individuals, representing a vast spectrum of opinion, participated and contributed to the development of a new way of confronting the subject of Government secrecy.
The data that we report here continue to support the need for reform. Classification decisions have decreased in recent years. However, in sheer nurnbers alone, the reproduction and distribution of classified information add many more millions of cla ssified pages to the classified universe each year. These additional pages far exceed the number that are declassified under the current system using available resources. With the likelihood of diminished personnel resources in the coming years, this tre nd will not change unless we adopt entirely new methods of classifying and declassifying information. At the same time, we cannot tolerate changes that undermine the national security. Seeking the right solutions within the context of these often competi ng circumstances has been and continues to be the goal of our efforts to restructure the security classification system.
April 26, 1993
PRESIDENTIAL REVIEW DIRECTIVE
SUBJECT: National Security Information
BACKGROUND - With the end of the Cold War, we should re-evaluate our security classification and safeguarding systems, as articulated in E.O. 12356, to ensure that they are in line with the reality of the current, rather than the past, threat potential.
OBJECTIVE - The objective of this tasking is to review E.O. 12356 and other directives relating to protection of national security information with a view toward drafting a new executive order that reflects the need to classify and safeguard national security information in the post Cold War period.
QUESTIONS - The following sets forth the questions that should be addressed durg this review. The resulting answers should serve as the basis for the drafting of the new proposed executive order which will be submitted upon completion of the review.
The Chairman of the task force shafl report to me through the NSC staff, Office of Intelligence Programs. The review should be completed no later than November 30, 1993, at which time a draft executive order superseding E.O. 12356 should be submitted for formal coordination.
PRD 29 establishes a 25-member interagencv task force to review the current classification system and prepare a draft replacement of E.O. 12356.
To build momentum and gain perspective, the task force begins by reviewing a vast amount of research compiled from a wide range of sources.
Individuals, including critics, from outside the Government are included in the initial stages of drafting an executive order on the classification system. More than a dozen people testify during two days of public hearings.
Over the following months, committees of the task force interview more than 100 persons, nearly half of whom are from outside the Government.
Members of the public and employees of the Government alike, submit over 100 documents to the task force that advocate everything from wholesale changes to maintaining the status quo.
As with the public hearings held earlier, divergent opinions on the classification system are again brought together as ISOO sponsors a Government-wide Declassification Conference.
A draft order with dozens of changes from the current system emerges.
On some issues, the chasm narrows between proponents of greater openness and proponents of greater secutity. Clearly gaps remain. All agree on the toughest questions that must ultimately be answered.
AUTOMATIC DECLASSIFICATION OF OLDER INFORMATION - Should classified informafion of permanent historical value be automatically declassified without further review when it reaches a specific age?
The symposium afforded many opportunities for attendees to interact with the guest speakers and each other. Many of the presentations included a question and answer period, which generated lively and sometimes controversial discussion. There was a consensus among the attendees that this type of forum provides opportunity to explore common concerns and new solutions, both of which are vital to the future declassification process.
R. Paul Richard
Deputy Staff Secretary to the President
Senior Research Analyst
Federation of American Scientists
Thomas S. Blanton
National Security Archive
Col. John A. Brown, USAF
Chief, Declassification Reference and Document Division
Defense Prisoner of War and Missing in Action Affairs Office
Office of the Secretary of Defense
Joseph H. Chaddic
Director, Historical Documents Review Division
Department of State
Sherry L. Davis
Chief, Document Classification Unit
Information Resources Division
Federal Bureau of Investigation
Paul R. Laplante
Chief, Policy Branch
Office of Declassification
Department of Energy
Larry M. Lawrence
Advanced Information Systems
Frank M. Machak
Office of Freedom of Information, Privacy and Classification Review
Department of State
Ella W. Nargele
Naval Historical Center
Department of the Navy
Thomas G. Paterson, Ph.D.
Professor, Department of History
University of Connecticut
Col Rodney Payne, USAF
Vice Commander of the Headquarters Air University
Maxwell Air Force Base
John E Pereira
Chief, Historical Review Group
Central Intelligence Agency
Mary I. Ronan
Chief , Access Staff
National Archives and Records Administration
Director, Records Declassification Division
National Archives and Records Administration
James J. Smith
National Security Agency
Air Force Declassification Panel
Lt. Col. Daniel J. Manix
Lt. Col. Don Dyrda
Maj. Dale E. Freeman
Col. Linda Smith
R. Paul Richard: There should be a commitment to openness. We should both herald our triumphs and reveal our mistakes so that our children and future generations are not doomed to repeat them. we should avoid routine classification ... and opt in favor of access.
R. Paul Richard: In the same spirit that we have examined our place in the world, so must we re-examine our role in the writing of that world's history. While ever vigilant for our security and foreign interests, we must keep in mind our greatest weapon: a knowledgeable citizenry that cares about and participates in its government.
Rodney Payne: In Air University's academic circle, we welcome your efforts [to declassiy more information]. For those efforts have already provided millions of pages which have been declassified and made available for study in all our schools through the Air Universitv Library and the Air Force Historical Research Agency. However, as an operator, I also understand that we cannot fight wars, if necessary, or win wars, if our techniques, tactics and procedures are compromised in trying to meet these demands for public openness.
Steven Aftergood: Government secrecy runs contrary to the way our political system is supposed to work. It tends to exclude the public and even, to some degree, their elected representatives, from some of the most important and consequential activities of government.
Paul Laplante: Openness is certainly a concept whose time has come to the RD [Restricted Data] world .... A lot of folks disagree on this openness initiative .... [T] his is going to happen; there's a lot of pressure out there for this to happen. Either we do it, or someone else is going to do it. And I guess it's better if we end up doing it - those ... with experience, and with the knowledge to do it right and in a balanced way.
Thomas Blanton: ... The open process [used in developing a new executive order] won't mean anything if the results aren't there, if you don't come up with a system that pretty immediately disgorges millions of real documents and a system that prevents the creation of new millions of documents for long term.
Frank Machak: [T] he most labor intensive and time-consuming aspect of the information access process is getting the final product to the requester. [T]he actual cutting and pasting and copying activity associated with redaction is particularly difficult from a manager's perspective in view of continuing limitations in human resources. It is becoming more apparent that the only approach to this activity can be through the application of technology.
Thomas Paterson: Create a clearinghouse in Washington that maintains computerized records of all documents declassified by any agency of the Government.
Jeanne Schauble: Vague [agency declassification] guidelines or lack of guidelines mean that we must withhold more material. And the more we must withhold, the slower the review process.
Joseph Chaddick: [W] e looked at new ways of doing things We put together a small team of foreign service specialists. In the four month period from June through the end of Fiscal Year 1993, they were able to complete the 1963 central records of the Department. It was over a million pages.
Paul Laplante: A very important point in any program along these lines is to get the highest level of commitment... The Secretary of the Energy, Hazel O'Leary, identified declassification as one of the six reinventing government initiatives within the Department of Energy...
John Pereira: ... We make the initial decisions. If other parts of the Agency disagree with us, then they have to be very specific about the damage to the national security. They have to be able to articulate what the specific damage would be, so mosaic is out. I've used mosaic for years in protecting information-- this piece and that piece go together and you have a story. That's out, unless you can show the specific lead-in for that mosaic .... We're, in effect, trying to change the mindset of CIA.
Ella Nargele: [T]here is a conflict between the [militarv] services. Every service has different [declassification] guidelines. There must be some means of resolving these conflicts between the services and between the agencies, too. Unless there is some method of doing this, we will be able to declassify less and less information in the joint operations environment that we face in the future.
Air Force Declassification Panel: [W]e thought that if we declassify as soon as possible after the event, we could stifl protect what needs to be protected and get the rest out the door and save the Air Force valuable storage space. [I]t allows us to protect ourselves against the critics and finally lets us reclaim our own Air Force history from a security file.
John Pereira: It's wise to get involved in the legislative process early ... If you have a choice between mandatory and voluntary, voluntary is best.
John Brown: The goal that the [Secretary of Defense] gave us [in response to the POW/MIA mandate] was to put the maximum amount of substantive information out on the streets. In fact, we have done that. Whether or not that stands the test of time remains to be seen.
Mary Ronan: ... I've seen the reductions. we're not talking about three words on a black page. We're talking about three words taken out of an entirely released page. So in that case ... agencies are in fact following the spirit of the bill and releasing an enormous amount of information.
Air Force Declassification Panel: In POW/MIA, we found out right away that many of these issues [problems associated with release of material] were legal issues, not declassification or classification issues.
Sherry Davis: One of the greatest impacts on our processing has been a reversal of a decision by the Attorney General on Communist Party-USA (CPUSA) matters .... It has ... had a positive effect on our processing of the JFK files, which is one of the leading cases of the day. The JFK CORE files and related files consist of over a million pages, at least two-thirds of which have been reviewed. Perhaps a third of this project will be released within the next two months.
For FY 1993, the number of original classifiers throughout the executive branch was 5,661, a decrease of 132 from the number reported last year. This figure represents the lowest number of original classifiers ever reported by ISOO. ISOO believes that efforts to downsize Government are largely responsible for the continued decreases in original classification authorities. There are disparities among agencies with comparable original classification atithority. Tlierefore, ISOO believes that additional reductions can be achieved witliout having a negative impact on agency operations. For this reason, ISOO will continue to urge agencies to keep the number of original classifiers at the lowest level possible to enhance the credibility of the classification svstem as a whole.
Another positive aspect of the decrease in the number of original classifiers is that all occurred at the Top Secret and Secret levels. ISOO believes that the benefits of reducing the number of original classifiers, especially at the higher levels, are significant. Limiting the number of original classifiers may result in fewer classification decisions. Because safeguards for documents classified Top Secret and Secret are more costly than those for Confidential, less original classifications at the higher levels should result in lower costs to the Government. In FY 1993, several agencies made special efforts to reduce the number of original classifiers and deserve special mention: ISOO applauds OMB and OVP for reporting decreases of 55% and 20%, respectively. DOD, DOE, Justice, NSC and USTR also reported considerable decreases.
For FY 1993, agencies reported a total of 245,951 original classification decisions. This figure represents a decrease of 49% over the original classification decisions reported in FY 1992, and is the lowest number of original classification decisions ever reported by ISOO. The main explanation for this dramatic decrease concerns changes in reporting by State. Prior to FY 1993, State routinely reported all classification decisions as original decisions. Because experience has shown that by far most classification decisions are derivative rather than origina1, ISOO had questioned this practice on many occasions. Finally, during FY 1993, State modified its reporting system to distinguish between original and derivative decisions. ISOO applauds State's commitment to continue to improve its information security practices. Though dramatic, the changes in State's reporting do not account for the entire decrease in FY 1993 original classifications. Seventeen percent of the decrease is attributed to other agencies' efforts to curtail original classification. By classification level, Top Secret classification decisions decreased by 14%; Secret by 49%; and Confidential by 52%.
Four agencies, DOD, Justice, CIA, and State, continue to account for almost 97 % of all original classification decisions. Of these agencies, DOD reported the highest number, with a total of 81,400 original classification decisions. Nevertheless, this number represents a significant decrease of 26% in original classifications at DOD. Even more impressive is the decrease of 40% in original decisions achieved by Justice. CIA also reported a 3% decrease in the number of their original classification decisions. For the agencies with smaller programs, the data collected show a 20% increase in the number of original classification decisions. While the increase is not a matter for serious concern, ISOO intends to monitor this trend closely.
As part of the original classification process, the classifier must determine a time frame for the protection of the information. This is commonly called "duration" of classification. E.O. 12356 provides classifiers with two means of designating declassification instructions for national security information. First, the information may be marked for declassification upon a specific date or event. For example, a classifier may determine that the information's sensitivity will lapse upon the completion of a particular project. The event would be noted on the face of the document, and when the project had been completed, the information would automatically be declassified. Only if a specific date or event cannot be determined at the time of classification does the classifier mark the document with the notation "Originating Agency's Determination Required" ("OADR"). "OADR" indicates that the information must be reviewed bv the originating agency before declassification action may be taken.
During FY 1993, only 3% of all original classification decisions were marked for declasification with a specific date or event, as compared to 5% of all original actions reported in FY 1992. Both of these proportions are too low. Changes to the classification system currently being considered, as directed by Presidential Review Directive (PRD) 29, include proposed solutions to the problem of indefinite duration of classification.
For FY 1993, the agencies reported 6,162,737 derivative classification actions.(*) This number represents a 5% increase from that reported in FY 1992. The increase is attributable to increases in derivative activity at both DOD and Justice/FBI, and the reporting by State of derivative actions for the first time. During FY 1993, State generated 153,373 derivative classification actions, 2% of the total of the three remaining agencies that account for almost 97% of derivative classification actions, Justice/FBI reported a 26% increase to 0.8 million, DOD reported a 10% increase to 3.6 million, and CIA reported an 18% decrease to 1.5 million.
As in the past, the breakdown of derivative classification actions bv classification level differs somewhat from the breakdown of orional decisions: Secret and Top Secret decisions continue to comprise higher percentages of the total. With respect to the proportion of Top Secret actions, this results from a very few activities that produce a relatively large quantity of derivative documents from classification guidance. Generally, this Top Secret information is highly localized, so that the percentage of Top Secret actions within almost all collections of classified information is much smaller.
While DOD accounted for 54% of all classification decisions reported for FY 1992, for FY 1993 it was 58%. CIA accounted for 25% of the total, Justice 13%, and State 3%. Again, the remaining agencies accounted for only 1% of the combined classification activity. These agencies run the gamut, however, in the degree of their involvement with classified information. They range from very large departments that possess very little classified information, and generate almost none, to very small entities that exist almost exclusively in a classified environment.
Both the number of pages reviewed and declassified decreased in FY 1993. Agencies reviewed 9 million pages, almost 1.6 million (16%) fewer than in FY 1992, and declassified 6.6 million pages, almost 3 million pages fewer. Agencies declassified 73% of the pages reviewed in FY 1993, a declassification rate significantly lower than the 88% declassified in FY 1992.
Although the decrease in systematic review is due to the low figures reported by several agencies, NARA's figure critically impacted on the systematic review product. In FY 1993, NARA reviewed 3 million pages, a decrease of over one-half the amount reviewed in FY 1992. Because the success of the systematic review program primarily rests with NARA, ISOO maintains a special interest in those matters that affect NARA's declassification program. NARA's explanation for the decline in its systematic review product consists of three factors: ( 1) the lack of resources to staff its systematic review responsibilities sufficiently; (2) NARA's required diversion of ten staff positions to fulfill the Congressional mandate concerning the Kennedy assassination files; and (3) NARA's move into the new Archives facility in College Park, Maryland.
For the fourth straight vear, Air Force accounted for most of DOD's systematic review activity. Of the 1,675,324 pages declassified by DOD, Air Force accounted for 56% of the total. Along with Air Force's performance, the efforts of State and CIA contributed significantly to the systematic review program for FY 1993. Both agencies significantly increased the number of pages reviewed and had declassification rates of 96% and 87%, respectively.
ISOO's annual reports have noted specific agency contributions to the systematic review effort; however, the agency-wide svstematic review effort has been described, in past reports, as "modest" and "disappointing." The voluntary nature of the program contributes to the non-participation by some agencies. ISOO continues to urge all agencies to contribute their fair share of effort to the systematic review program, even as the PRD process searches for more effective ways of dealing with the huge backlog of older classified records.
The 3,911 cases processed under mandatory review during FY 1993 comprised 36,157 documents totaling 246,903 pages. The number of pages processed represents a 6% increase from the prior year. The percentage of pages declassified in whole or in part (93%) remained at almost the same high level as last year's rate of 94%. Given the high proportion and number of pages declassified, mandatory review remains a highly successful mechanism for the declassification of information.
E.O. 12356 also provides that agencies or members of the public may appeal mandatory review denials to designated officials of the denying agencies. During FY 1993, agencies processed 242 appeals that comprised 12,104 documents totaling 38,094 pages. Of these, 97% of the pages were granted in whole or in part. This impressive rate suggests that researchers can anticipate even greater retums in declassified information if they pursue an appeal.
For the third year in a row, agencies reported a decrease in the number of self-inspections. For FY 1993, agencies reported 1,967 fewer self-inspections, a 9% decrease in the number reported in FY 1992. This significant decrease is largely attributed to DOD, which conducted 1,800 fewer self-inspections in FY 1993 than in FY 1992. DOD's reduction in self-inspections can be attributed to downsizing and reorganizations within the services. Other agencies with significant decreases include CIA, DOT, State and NRC. Those agencies reporting major increases, thus enhancing their oversight capability include Commerce, DOE and Justice.
In FY 1993, agencies detected a total of 18,165 infractions. Compared to FY 1992, this figure represents an 11% decrease. Although the overall number of inspections has decreased by a substantial margin, the average number of infractions discovered per inspection decreased slightly: from 0.99 in FY 1992, to 0.97 in FY 1993. This rate indicates that not all agencies leave effective self-inspection programs. ISOO has consistently held that agencies would identify a far greater number of infractions if agencies conducted more quality self-inspections. If agencies have not already incorporated periodic samplings of their classified product into their self-inspection programs, ISOO strongly encourages them to do so. Although the overall number of infractions decreased, an increase in the mismarking category appears to indicate that more agencies are incorporating document reviews as part of their self-inspection programs. ISOO applauds this and hopes that this trend will continue.