August 25, 1997
Dr. Richard A. Meserve, Chairman
Partner, Covington & Burling, Washington, D.C.
Mr. David H. Albright
President, Institute for Science and International Security, Washington, D.C.
Ms. Cheryl Alexander
President, Alexander Companies, Wayzata, Minnesota
Dr. Thomas A. Cotton
Vice President, JK Research Associates, Inc., Vienna, Virginia
Dr. Douglas M. Eardley
Professor, Institute for Theoretical Physics, University of California, Santa Barbara, California
Dr. E.G. Mahler
President, E.G. Mahler & Associates, Inc., Wilmington, Delaware
Dr. Page P. Miller
Director, National Coordinating Committee for the Promotion of History, Washington, D.C.
Dr. Al Narath
President, Energy and Environment Sector, Lockheed Martin Corporation,
Albuquerque, New Mexico
Dr. Allen Lee Sessoms
President, Queens College, The City University of New York, Flushing, New York
Mr. Troy E. Wade, II
President, Wade Associates, Inc., Las Vegas, Nevada
Ms. Ellyn R. Weiss
Partner, Foley, Hoag and Eliot, Washington, D.C.
Dr. Eric H. Willis
Counselor, DynMeridian Corporation, Arlington, Virginia
By “responsible openness” we mean both a mindset and a set of policies by which the Department recognizes and affirmatively seeks to fulfill its obligations to provide the public with accurate and complete information about its activities to the maximum extent consistent with protection of national security and with other societal objectives (for example, protection of personal privacy). We believe that such openness is an essential precondition — in fact, an imperative — for the Department if it is to achieve its objectives. In this report, we shall seek to establish the reasons for this viewpoint and to offer some recommendations as to how improvements can be achieved.
We have been significantly assisted in our work by the several comprehensive studies of classification issues. The Commission on Protecting and Reducing Government Secrecy (the Moynihan Commission) has recently issued a very thoughtful report that urges a systematic and government-wide rethinking of secrecy issues.1 We have had the benefit of the reviews of the Depart-ment’s classification system by the National Academy of Sciences.2 And, the Department has itself engaged in a careful examination of classification issues, most recently in a comprehensive policy review.3 Our burden has thus been considerably lessened by the efforts of others.
We also hasten to add that the Department has made very significant strides in recent years in seeking to achieve openness — progress for which the Department should be commended. These efforts have included the declassification and release of a significant volume of information of policy or public interest, including information about the nuclear stockpile, nuclear tests, environmental releases, human radiation experiments, and a variety of other matters. The Fundamental Classification Policy Review, an effort that is now near completion, reflects the continuing efforts by the Department to improve its policies and practices. While the Panel has thus had the challenge of seeking to comment on a system that is changing as we are conducting our work, we have the benefit of a Department that is poised to respond.
We have been ably assisted in our work by the SEAB staff and the Office of Declassification. We very much appreciate their contributions.
Richard A. Meserve
In order to reform the classification system, the Secretary should build on the significant work that has been undertaken in the Fundamental Classification Policy Review and should implement the recommendations arising from that effort. For those items of decontrol on which interagency agreement has been reached, the Department should proceed promptly with the revision of its classification guides. Moreover, the Department, at an appropriate time, should support the amendment of the Atomic Energy Act so as to establish that an affirmative action by Government is necessary to classify information. The restrictions on Unclassified Controlled Nuclear Information (UCNI) should be confined to safeguards and security information and restrictions even in this limited area should be subject to careful examination to determine whether controls are necessary and, if so, whether classification is a preferred means of providing protection.
In order to reduce the burden of the classification system in the future, the Department should encourage practices that minimize classification. For example, DOE should follow the procedures of certain other departments and, where possible, documents should be unclassified or have their classified content reserved to a classified annex. The classified portions of documents should be clearly and separately marked.
The Department has developed certain “finding aids” — guides that can help locate documents of possible interest — in connection with various targeted document reviews (for example, efforts to find records relating to human radiation exposure). In order to improve the system, the Department should seek to extract the lessons that can be learned from these past efforts and should compile a centralized directory of the currently available finding aids. It should also expand its efforts by developing a uniform format for new finding aids and should experiment with such aids in important topical areas, such as documents relating to the evolution of radiation protection standards or to fissile material production.
The Department should also seek to enhance efficiency through the use of technology. There are a number of technologies, some of which are in use by the Department, for scanning paper documents and saving them both as images and as text files that are subject to rapid computerized searches. The Department should explore the feasibility of a broad-scale document management system that is tailored to the Department’s needs and to fund one or more pilot tests. The Department should also explore the use of artificial intelligence to facilitate declassification reviews, even though human review is likely to be necessary for the foreseeable future.
Substantial amounts of information under the control of DOE are now maintained in electronic format as word-processing documents, databases, and e-mail. The shift from paper records entails a significant reorientation of the procedures for the maintenance of information. (Indeed, electronic records may be even more perishable than the paper records that they replace.) The challenge presented by electronic records must be addressed urgently on a Department-wide, if not a Government-wide, basis. While the development of uniform standards for preservation and access may be difficult to attain, a failure to address this problem now will allow today’s confusion to develop into tomorrow’s chaos. The Department, with the high level of computational proficiency in the national laboratories, may be particularly well suited to taking an active role in addressing this urgent problem.
Finally, the permanence of any gains will be threatened unless changes in the old ways of doing business are seen by all to be in their self-interest. Openness should be established as a core value of the Department through incorporation in performance reviews, program plans and contracting activities. For the foreseeable future, openness requires sustained resources and continuing Secretarial attention and emphasis.
Not surprisingly, the many decades of secrecy stemming from the Department’s stewardship of the nuclear weapons complex and its critical role in creating a formidable nuclear arsenal served to establish a “culture” of secrecy. However, now that the Cold War is over and the threats have changed substantially, it is time to reassess the costs and benefits of that culture, and consider whether changes are required to meet the challenges of the future.
DOE needs to have the public trust if it is to accomplish its missions. The Department’s capacity to fulfill its missions in national security and other areas ultimately must derive from confidence in the Department by the American public and their elected officials. Excessive secrecy — and the suspicions that it can encourage — can have a corrosive influence on public attitudes toward the Department.
In coming years, DOE must carry out major responsibilities involving nuclear weapons, the management of radioactive materials, and environmental remediation. These will require DOE to select new facilities for producing radioactive tritium (so as to maintain the viability of the current inventory of nuclear weapons), and other facilities for processing radioactive wastes and surplus fissile materials and for storing and finally disposing of these materials. In addition, DOE will have to transport radioactive materials through many communities throughout the country. Given the high level of public concern and sensitivity about radioactive materials, and the continuing debate about nuclear weapons and nuclear power, these would be challenging tasks in the best of circumstances. The difficulty will be aggravated if the Department is suspected of hiding risks and of concealing past accidents. Openness — and the enhanced credibility that can come from it — is a necessary condition for success in these activities.
Greater openness may also be required for the successful achievement of the Department’s missions for another, more subtle reason. The Department faces a critical time of transition to an environment of no weapons testing and the possibility of accompanying structural changes within the weapons complex. DOE confronts a major challenge as it pursues an agency mission — the assurance of a credible and reliable nuclear deterrent — that may seem increasingly anachronistic with the end of the Cold War and the disappearance of our major military rival. This reexamination has its greatest programmatic thrust in the program for “Science Based Stockpile Stewardship” (SBSS).
The success of the SBSS program will depend in part on the Department’s ability to recruit and retain a staff of highly skilled scientific and technical professionals. But, a life “behind the fence” may not seem as desirable to new recruits as it may have been during the Cold War. Indeed, with the major DOE labs moving toward new missions and ways of operating, the closed lifestyle of the past may prove increasingly difficult to maintain. Finding ways to assure some openness in the work — or at least efforts to avoid needless secrecy — may prove essential in recruiting a cadre of talented new scientists and engineers to replace those who are moving into retirement or other fields.
Moreover, the restructuring of the weapons program will present some special challenges. SBSS will involve activities that overlap with unclassified work: simulation using advanced computing technologies will replace testing, and experiments will be performed on dual-use machines such as the National Ignition Facility. The productivity of the laboratories will thus probably entail a greater mix of classified and unclassified research than in the past. The less classification there is, consistent with rigorously protecting highly sensitive information, the more conducive the climate is likely to be to productive advance.
The Department is already launched on a path to greater openness. The Department has wisely concluded that its best response to the decay of public confidence is a candid and forthright willingness to acknowledge and confront its past. Through the Openness Initiative launched in 1993, DOE has committed itself to providing the public with records and information. Indeed, some of the most difficult and controversial steps in exposing a hitherto hidden past to public scrutiny have already been taken. In addition, in many areas the Department has also begun to conduct its business in a much more open manner than has been the past practice, with substantial involvement of interested and affected parties in the programs of concern to them. These steps have been well received by the public and various stakeholders, and DOE deserves credit for achieving so much in a short period.
However, much remains to be done to institutionalize and to expand the improvements that have been made to date. Openness should be a normal part of doing business in the Department, sustained by high-level management attention, but not requiring continued high-level prodding. The challenge facing the Department today is to convert openness from a new initiative to a standard operating procedure. Moreover, the Department must find the means to sustain the effort at a time when Congress has reduced the funds for this important activity.
The Secretary should place a high priority on enhancing and institutionalizing openness throughout DOE and its contractor community. The public trust that openness can nurture is an essential precondition for success in the Department’s activities.
We hasten to add, that while openness does mean declassification of information that no longer requires classification — as well as steps to ensure public access to documents containing information that is unclassified or has been declassified — it most emphatically does not mean the release of information that should be safeguarded. There is good reason that Restricted Data concerning nuclear weapons has been subjected to a statutory classification and protection system. Information about nuclear-weapon design does not lose its value to a potential adversary with the passage of time. Indeed, some of the oldest (and thus least technically sophisticated) nuclear-weapon designs may be of the greatest interest to potential nuclear proliferators. Responsible openness thus entails retaining careful protection of certain types of information. But it also involves defining a boundary for the scope of the classification that is different from that established in the past. This and related issues are discussed further in Chapter 1.
In June 1991, members of an International Atomic Energy Agency (IAEA) Inspection Team were denied access to a military site near Al Fallujah, Iraq. Inspectors saw dozens of Iraqi trucks loaded with machinery and equipment leaving the site by a back gate. They followed the trucks in their own vehicles and learned that some of the trucks carried 12-foot diameter, 60-ton magnets. They were forced to give up the chase after warning shots were fired, but nevertheless obtained photographic evidence of a uranium enrichment program based on electromagnetic isotope separation, a technology long considered obsolete by the nuclear weapons states.
Electromagnetic isotope separation (EMIS) devices called “calutrons” were used in the Manhattan Project and helped produce the enriched uranium in the bomb dropped at Hiroshima. They are based on the principle that as charged particles pass through a magnetic field, heavier particles will be deflected less than lighter ones with the same charge and kinetic energy. This enables the separation of different isotopes of uranium.
For more than 40 years, however, the nuclear powers have used more efficient technologies to enrich uranium, in particular gaseous diffusion and gas centrifuge processes. Most experts believed that would-be proliferators would choose one of these technical approaches.
According to the IAEA Action Team, which is charged by the United Nations Security Council with conducting nuclear inspections in Iraq, the Iraqi separator design was based on 1940s U.S. calutron design. Iraq had decided in the early 1980s, when it formally launched its enrichment program, that EMIS was the most suitable technology for its inexperienced scientists and engineers who had to design and manufacture the EMIS components.
Iraq sought to improve the U.S. design with mixed results. In the case of certain key EMIS components, Iraq’s extensive attempts to improve the design failed, leading Iraqi scientists to revert back to the original U.S. design. Nonetheless, because some improvements were made, Iraq prefers to call its machine a “Baghdadtron.”
This example illustrates that technologies considered obsolete by the United States may still be useful to proliferators and require careful review before declassification.
Sources: DOE/Office of Arms Control and Nonproliferation, Critical Technologies Newsletter, Volume 10, Issue 3 (December 1992); International Atomic Energy Agency Action Team reports.
Even if all of the information in DOE’s possession were declassified immediately, openness would still not be achieved. For example, much of the justification for the Openness Initiative came from public interest in the environmental consequences of activities in the Department’s weapons complex and in the studies on health effects from radiation exposures. Virtually all information bearing on environment, health, and safety is now unclassified. But, the simple fact that the information is unclassified does not necessarily mean that it is accessible. Unclassified information that is buried in a file is effectively unavailable to the public (or the Department). Moreover, many of the documents embodying such information may also contain classified material and therefore have not yet been publicly released. Ultimate achievement of openness thus requires identification of relevant materials in the huge collection of documents under the Department’s “control” and the public release of appropriate documents (or the information they contain). Increasing openness thus presents a difficult document-examination, document-control, and communication problem.
Historically poor record keeping in the Department of Energy and its predecessors, compounded by decentralized, contractor-managed and production-driven operations, has led to a situation in which the Department literally does not know in many cases what records it has or where to find them. This is exacerbated by the policies of secrecy entrenched in the Cold War environment. This combination of factors means that even when records are not specifically classified, the Department often lacks effective and credible mechanisms to make them accessible. Moreover, the Department should put in place procedures to assure accessibility to information that is now being generated or that is created in the future. Approaches to this problem are discussed in Chapter 2.
Until cultural change is seen by all to be in the self interest of the Department’s and its contractors’ employees, lasting and fundamental changes in the way DOE does business will be difficult to achieve, and the advances of the last few years will be transitory achievements. This problem is discussed in Chapter 3.
It is difficult, however, to define clearly the full range of information that should be protected. All agree that non-public information on the construction of a nuclear weapon should be controlled, but there are legitimate questions as to how far the veil of secrecy should extend. Should it include information about past nuclear tests? Should it include information on the environmental or health impacts from weapons-related operations? Should it include information on technologies that, while bearing on weapons production, have uses in civilian products or manufacture?
Developing a sound policy on classification requires the balancing of overlapping, reinforcing, and at times competing considerations. The main considerations fall into several general categories. Only by weighing the relevant factors can a proper balance be achieved.
The narrowing of the information to be protected enables the focusing of resources — time, money, effort — thereby providing greater assurance that core information is truly safeguarded. Diluting resources by sweeping too much under guard may lessen the protection of the “crown jewels.” Indeed, staff and contractors may underestimate the importance of stringent adherence to classification rules if they perceive that the system encompasses information of trivial national security significance.
Moreover, the direct financial costs of the classification system are only a small portion of the true overall costs. Classification — with its burdensome restraints on access — no doubt significantly limits productivity. And, by inhibiting the cross-fertilization of ideas that openness can encourage, classification can constrain technical advance. To the extent that classification imposes limits on scientific or technical exchange that could bear fruit in the civilian economy, it restrains our overall economic advancement and growth. Consideration of efficiency thus argues for narrowing the purview of the classification system.
Moreover, considerable nuclear information is already publicly available. As a result, knowledgeable observers have stated that obtaining classified information may not be necessary for a potential proliferator to construct a weapon, although of course such access may greatly facilitate his work or increase the reliability and efficiency of the resulting device.5 Classification is an important tool in preventing or slowing down nuclear proliferation, but it alone cannot do the job.
Openness on the part of the United States can encourage reciprocal openness by other nations, reducing tensions and enhancing national security. In late 1994, the Department of Energy declassified and published the remaining secret information regarding the dates and yields of U.S. tests. Secretary O’Leary called upon her Russian counterpart to make a comparable public accounting, stating that the “release of this information should also encourage other nuclear weapon nations to declassify similar information.”
Russia reciprocated in July 1996, when Victor Mikhailov, Russian Minister of Atomic Energy, presented Secretary O’Leary with the first-ever report on U.S.S.R. Nuclear Weapons Tests and Peaceful Nuclear Explosions — 1949 through 1990. In the foreword of that report, Mikhailov cited the U.S. publication and noted the Russian report’s symmetry with the earlier DOE report.
Source: U.S. Department of Energy, United States Nuclear Tests July 1945 through September 1992, Report DOE/NV–209 (Rev. 14), Washington, DC: U.S. Department of Energy, December 1994.
This Panel has not had the opportunity, nor does it have the expertise, to reexamine the specific technical conclusions of the review group. However, we are impressed by their effort. We recommend prompt action on the review group’s recommendations.
We understand that achieving final consensus on the review group’s recommendations has not yet been achieved. In order to maintain the momentum of the review group’s efforts, the Department should take the following steps:
In order to bring about an actual change in classification practices, it is necessary to implement the policy recommendations through modification of the classification guides that provide concrete instructions on classification matters. This is a substantial task: there are more than 50 headquarters classification guides and about 800 local guides. Because the benefits of a revised classification policy cannot be achieved until the classification guides are modified, the effort should start now.
Different analysts will weigh the costs and benefits of openness in different ways and thus some interagency friction on declassification matters is to be expected. Given the extensive work that has already been undertaken by the review group to provide the factual foundation for discussion, the Department should pursue efforts to achieve closure with respect to the items on which interagency agreement has been difficult to achieve. If the final resolution of an item is one that endorses the review group’s recommendation, then the Department should press forward with the associated modification of the classification guides.
While the rule does not concern itself primarily with what information is to be classified, it does provide detailed guidance as to how classification decisions will be made. The Department’s efforts to develop this guidance by way of notice-and-comment rulemaking represents a radical departure from the past practice of simply imposing requirements by departmental order without any explicit opportunity for public involvement or judicial review. The Department’s decision to proceed by way of notice-and-comment rulemaking should be seen to represent a corollary activity to expanded openness: the Department is also adopting its procedures through open processes. We commend the Department for its classification rulemaking.
Until recently, this accumulation was still growing, as new classified documents were being created faster than old ones were being declassified. According to the Department, the corner was turned in 1996 when the number of newly generated classified documents fell below the number of documents that were declassified. Nonetheless, a huge backlog remains.
The Department’s declassification efforts are driven by a large number of external demands, many of which (for example, court orders or congressional requests) simply cannot be postponed. As a result, DOE has relatively little discretion to “set priorities” for declassification of documents. This lack of flexibility has been compounded by recent budget cuts, which have reduced the number of personnel available for declassification efforts. There are now 26 authorized RD declassifiers at Headquarters, down from 35, and 18 trained for National Security Information reviews, down from 35. Instability in funding is a serious problem because it undermines the Department’s capacity to maintain a cadre of skilled personnel. And budget pressures can also be expected to affect adversely the priority attached to declassification efforts at the DOE field sites.
Percent of Total Records Classified 5% Unclassified 95% Location of Records DOE Offices 15% (headquarters and field) Laboratories 65% Facilities, Contractors, and 20% OtherNote: The data presented are approximate.
In view of the limitations in resources, it is appropriate that declassification efforts be driven by demand from potential users. DOE should therefore focus on informing that demand by providing systematic information about what documents exist, what they contain (in general or topical terms), whether they are classified, and where they can be located. This will allow interested parties to target their requests with some specificity, and could well lead to a more efficient use of DOE’s limited resources. Means to accomplish this objective for both classified and unclassified documents are discussed in the next chapter.
In 1994, Congress established the Department of Energy’s Declassification Productivity Initiative (DPI). The initiative’s goal is to improve the efficiency and accuracy of the document declassification review process by creating computer automation tools. Because of the extreme sensitivity of the documents under review, two human reviews are now required to minimize the possibility of error.
DPI’s ultimate goal is to produce a computer program that can do the first of the two declassification reviews now done by highly skilled personnel. This task is not a simple one and will require both sustained effort and support to bring it to fruition. The strategy is to make incremental improvements in the program to make more efficient use of the human reviewers’ expertise and time and thereby increase their productivity. This effort, the first of its kind in government, is also expected to have wide application in other government departments.
Source: U.S. Department of Energy, Office of Declassification.
The Department should encourage practices that minimize classification. Where possible, documents should be unclassified or have their classified content reserved to a classified annex. The classified portion of a document should also be clearly and separately marked. The current approach of mixing classified and unclassified information requires that the whole content of each classified document must be protected pending declassification review. Separating classified information from unclassified information at the time of the creation of documents should significantly reduce the cost of record stewardship in the future.
Other departments and agencies, in particular the Department of Defense, have already begun the practice of producing unclassified versions of reports, particularly those of widespread public interest. When necessary, these reports have classified appendices that can be stored separately. This “dual track” approach permits more information to be made available to the public immediately. It also radically reduces the need for classification review.
Modern computer capability should be harnessed to assist in the classification process for future documents. For example, all documents at their creation, whether classified or not, might be given a precisely defined computer designation. This “tag” might include the date, subject area, topic area, type of document (for example, memorandum, policy paper, technical report) and other indicators deemed useful by potential users. (The “tags” should be uniform throughout the DOE complex.) The tags could also serve as “finding aids” for researchers, both within and outside the Federal Government, and should greatly facilitate document management. Further discussion of the use of computer technology to facilitate access to and control of all DOE documents is found in the next chapter.
Unfortunately, the Department of Energy has neither ready access to the documents under its control nor an adequate inventory of those documents. As noted above, according to its own 1995 estimate, the Department is the steward of approximately 2.7 million cubic feet (roughly 6.75 billion pages) of documents in many forms (from laboratory notebooks to policy memoranda) of which it has, at best, incomplete knowledge. Thus, means to address records accessibility overlap with classification issues, but extend far beyond classification.
DOE estimates that more than 90 percent of its records have been inventoried. This means that the records have been described adequately to develop a records disposition schedule (the legal authority by which federal records may be destroyed, transferred, or otherwise alienated from agency custody). An inventory is conducted at the series level — that is, a level at which records are kept together because they relate to a specific activity, function, or subject. In making an inventory, one goes into an office, notes descriptive information on file drawers or file folders, does a quick survey to ensure the consistency of that descriptive information to the drawer or folder contents, notes the range of dates for the records, and calculates volume. Only general information, sufficient to develop a records disposition schedule, is collected in a records inventory. An inventory does not provide an itemized index of specific documents and thus does not enable one to find a specific document quickly. Drawing an analogy to a grocery store inventory, such an inventory could reveal that the store has 10,000 cans of soup, but would not tell you where to find a specific can of Campbell’s soup.
Source: Data from the DOE Office of Records Management and the Office of Declassification.
The current situation — a vast, poorly understood accumulation of classified and unclassified documents under little or no control — is not merely an offense against good recordkeeping practices. It can have serious legal and financial consequences for the Department (see “The Price of Neglect” opposite [below]).
This situation has made it difficult and costly for the Department to meet its obligations in litigation, in responding to Freedom of Information Act (FOIA) requests, in preparing congressional testimony, and in satisfying other demands for historical information. It has also led to embarrassment and needless lawsuits and even the threat of judicial sanctions to punish DOE for lack of responsiveness, and incomplete responses, to requests for information. Indeed, DOE’s inability to access its own documents means that the Department has limited memory of its own past actions, which can frustrate its capacity to achieve its current missions in an efficient fashion. Moreover, the inability to locate relevant documents also feeds the public’s suspicions that something sinister is being hidden from them.
At a more immediate level, the lack of control of historical records makes it arduous and time-consuming for the Department to retrieve information from its own files. For example, the collection of information concerning human radiation experiments was difficult because of poor record management practices, not because of classification (most of the relevant documents were not classified). If the Department hopes to respond more quickly to demands for information in the future, steps must be taken in advance to gain better intellectual control of its records.
The Department must improve its document control systems. Through the Openness Initiative, DOE has committed to providing the public with records and information and, over the last several years, DOE has taken many steps to facilitate such public access. However, much remains to be done. Until DOE has better control of its records, it cannot fully realize its openness goals. In implementing the Information Technology Management Reform Act of 1996, 11 DOE should ensure that the Chief Information Officer has the authority and resources to improve document-control and records-management practices across all elements of the Department. The Chief Information Officer should be given the authority and resources to address both paper and electronic records, as discussed below.
In a still-unresolved lawsuit alleging damages caused by releases of radioactive materials from DOE’s Rocky Flats facility, the Department’s lack of control of documents led to a contempt order against DOE for failure to comply with a stipulated agreement to produce documents requested by the plaintiffs.
When DOE attempted to comply with a document production schedule, it made several unpleasant discoveries. In December 1995, DOE estimated that the plaintiffs’ request for documents related to materials unaccounted for (MUF) would require declassification review of 11,000 pages. In January 1996, the estimate was increased to 400,000 to 500,000 pages, and in February to 670,000 pages. At about the same time, the Rocky Flats contractor located 1,500 reels of microfilm containing documents.
All of this led the plaintiffs to charge that the Department had been making statements about the extent of information in its possession that it knew, or should have known, to be false.
When DOE recognized the true magnitude of the declassification task, it realized that it could not comply with the declassification plan and procedures it had previously accepted in a 1994 stipulated order, and could not meet the court-ordered deadline for completing the declassification review, despite a doubling of the declassification staff from 14 to 28. Because of the great suspicion that was created, the plaintiffs were unwilling to accept DOE’s assertions that the MUF documents contained so much highly sensitive information that, when they were reviewed and the classified portions redacted, little useful information would remain.
DOE was forced to divert substantial Headquarters and field resources to a large-scale review of documents. Ultimately, the plaintiffs agreed that the effort was largely a waste of resources because of the low relevance of the small amount of information that could be declassified. Millions of dollars were wasted and DOE’s credibility was damaged. And, since the case is not yet closed, the possibility of further embarrassment exists.
Availability of good finding aids has the potential to reduce substantially the cost of searches for documents. A report on the impacts of the Openness Initiative on one field office noted that while a search for a known document with a known title costs only around $200, a search in response to a broad request for all information concerning a general topic could come to nearly $14,000. 13 Aids that enable someone seeking a document to increase the specificity of his or her request can reduce the costs of a response.
There are several steps the Department could take to continue to develop and disseminate finding aids to the huge inventory of documents in its possession:
In response to a recommendation by the Presidential Advisory Committee on Human Radiation Experiments, the DOE Office of Human Radiation Experiments (OHRE) has already undertaken a project aimed at making finding aids to inactive records in DOE custody available to the public. In this project, OHRE has collected lists of folder titles for record series of potential interest from across the DOE, added brief introductions that provide background and context for the series, and placed the listings and introductions in public reading rooms. This information will soon be added to the OHRE site on the World Wide Web. This effort should be continued and expanded to include all currently available finding aids that exist for the Department’s records. The information should be placed directly in OpenNet and should extend beyond health- and safety-related topics. The Panel strongly encourages DOE field sites to cooperate with OHRE (which has been renamed the Office of Research, Records, Data and Access) in this effort.
The review yields only pass/fail decisions: there will be no effort to redact classified documents so that the unclassified portions are made public. Nonetheless, the review offers an opportunity to create finding aids to classified documents that can be valuable in later declassification efforts by enhancing the specificity and focus of subsequent requests for classification reviews. The National Academy of Sciences committee cited experience during the dose-reconstruction project at Hanford that showed that even a simple list of titles of a classified document was a substantial help in focusing the search for information.
DOE is already acting to place index information about declassified documents on OpenNet, and to forward the declassified documents for placement into public reading rooms. To facilitate demand-driven declassification, existing and new finding aids to still-classified documents should be disseminated through OpenNet, DOE reading rooms, and other means.
The shift to electronic media has fundamental implications for records management. While the paper-based domain had been essentially static for years, the new age of electronic information technology is advancing at a prodigious pace. The amazing technological advances in hardware are being matched by increasingly sophisticated software. Indeed, it is not unreasonable to suppose that the advance of hardware and software will continue apace or even accelerate, and will develop in ways that are difficult to predict today.
As a consequence, electronic media may be far more “perishable” than the paper they replace. Who remembers the eight-track tapes in the seventies, the Beta videotapes of the eighties, or the punch tapes and punch cards of early computers? The computer revolution thus presents a challenge for those who seek to safeguard information generated in soon-to-be-archaic formats. We have no ready answer to this problem, but for the foreseeable future, electronic document management systems must at the least be designed with sufficient flexibility so as to adapt to this constant change.
In 1985, the Committee on Records of Government, a blue-ribbon panel created by the American Council of Learned Societies, the Council on Library Resources, and the Social Science Research Council, released a report on government records. Among its principal conclusions were:
The danger of losing historically valuable records is greatly increased by the changeover to electronic recordkeeping. Under current procedure, records created on tapes or disks are erased or lost before anyone exercises judgement about their possible value. In addition, given the rapidity of technological change, even information recognized as valuable can be lost because the equipment and skills necessary to retrieve it become obsolete or unavailable.
They also noted that:
By the mid 1970s, when computer tapes for the 1960 census came to the attention of archivists, there remained only two machines capable of reading them. One was already in the Smithsonian. The other was in Japan!
Because of erasure of electronic records, future historians may know less about the Reagan Administration’s 1985 arms control initiatives than about those of 1972 which led to SALT I or, for that matter, those of 1921 which led to the Washington Naval treaties.
Even without the problems posed by rapid technological advance, the electronic revolution may make the future research into the foundations for policy more difficult. In the electronic domain, there may be little or no “paper trail” to facilitate interpretation. For example, the use of a word processor enables a document to evolve without necessarily leaving a record of changes. Comments by reviewers are incorporated electronically into a new draft, the earlier draft is effectively erased, and there is no record of the “debate” that went into the final product. Often it is this debate, captured in the past in the working documents leading up to the final version, that evokes the interest of the historian, or which turns out to be germane to legal and congressional proceedings. With the shift to electronic media, the whole process by which decisions were reached, and the diversity of the views that went into them, are in danger of being lost forever. Persons 20 years from now faced with examining remaining records for historical or litigation purposes may have no insight as to how decisions, critical and mundane alike, were reached.
The implications of these dramatic changes have yet to draw policymakers’ attention to the need to control the generation, storage, and retrieval of information in the future in a way that is fundamentally different from the familiar paper regime. We are entering uncharted territory, seemingly preoccupied with gazing into the rear-view mirror at the past, rather than through the windshield into the future. The Federal Government as a whole — not just the Department of Energy — appears unprepared for the new challenge. A Justice Department attorney recently was quoted as saying that “[w]hen it comes to preserving computer records in an electronic format, the vast majority of government agencies simply are not equipped to do that.”15
The Panel believes that to ensure openness in the future, the challenge presented by electronic records must be addressed urgently on a Department-wide, if not a Government-wide, basis. There is an awareness of the problem in interagency circles, and some agencies are seeking to achieve standardization in the management of electronic records. Nonetheless, the problem is so pressing that DOE may not be able to afford to await development of a Government-wide consensus on uniform standards for the generation, storage, archiving, and retrieval of electronic information. While experience suggests that uniform standards may be difficult to attain, a failure to address this problem now will allow today’s confusion to develop into tomorrow’s chaos.
The Department possesses a unique resource to apply to the problem unavailable to any other agency of government: the national laboratories, with their high level of computational proficiency, may be well suited to taking an active role in addressing the integrity of future recordkeeping in the face of continuing technological change. Indeed, this capability might enable DOE to take the lead in this challenging task.
As discussed in the Introduction, the achievement of the Department’s missions, including especially its critical mission concerning nuclear weapons, requires significant efforts to change the ingrained habits of 50 years of secrecy. Changes in policy will be ineffective unless the culture of the Department is also changed.
The achievement of change is complicated by the fact that there is no single focal point or clearly defined budget for declassification and document-management efforts within DOE. The roles and responsibilities are divided between DOE Headquarters and the DOE operations offices and management and operating (M&O) contractors. The primary role of Headquarters is to establish policy and guidance. But the accomplishment of the work must rest principally with the operations offices and the contractors.
Operations office managers have significant authority over their respective budgets and staff, and decide how much to spend on classification and declassification activities and on document management at each site. Usually such activities are funded as overhead costs associated with other activities. Because the operations offices are funded primarily through DOE’s program offices (primarily Defense Programs and Environmental Management), they understandably respond to the priorities of those offices. Viewed in this way, a Headquarters directive to perform a classification review or to upgrade document management is essentially an “unfunded mandate” whose execution depends on the willingness of the field offices to allocate the necessary resources. The result is an inconsistently implemented program. In the case of the Large-Scale Review, for example, most field offices have submitted plans for approval, but only a few have found the funding to begin the review in earnest.16
The diffusion of responsibility and lack of central control is further compounded by the fact that in many cases documents are in the possession of DOE contractors, who have considerable discretion about whether to fund efforts to review documents for declassification and to make declassified or unclassified documents available for outsiders to use. The Department should undertake a systematic examination of its management of classification and document-control activities. The Department should find the means for ensuring that classification and document-control issues receive appropriate resources and attention by those who have responsibility for implementing revised policy.
Finally, until changes in the old ways of doing business are seen to be in the Department’s, the employees’ and the contractors’ self-interest, lasting and fundamental changes will not occur. Indeed, permanence of the gains that have been made will be threatened. In order to achieve and maintain change, the following steps should be undertaken:
2. National Academy of Sciences, A Review of the Department of Energy Classification Policy and Practice (1995). The National Academy of Sciences also issued a second report that responded to the Department’s request that the Academy examine the adequacy of the Department’s response to the initial report: National Academy of Sciences, Review of the Department of Energy’s Response to the Recommendations in the National Research Council Study of DOE Declassification Policy and Practice (July 1996). Certain members of this advisory panel (Meserve, Cotton) participated in the Academy’s efforts.
3. DOE Fundamental Classification Review Group, Fundamental Classification Policy Review (Draft, Feb. 1, 1996). Also, R. Pollock, et. al., Classification Policy Study, (July 1992).
4. National Academy of Sciences, A Review of the Department of Energy Classification Policy and Practice (1995); Review of the Department of Energy’s Response to the Recommendations in the National Research Council Study of DOE Declassification Policy and Practice (July, 1996); and Moynihan, et. al., Commission on Protecting and Reducing Government Secrecy, Secrecy: Report of the Commission on Protecting and Reducing Government Secrecy (1997).
5. National Academy of Sciences, Management and Disposition of Excess Weapons Plutonium (1994).
6. Narath, et. al., Report of the Fundamental Classification Policy Review Group (1996).
7. These working groups examined weapon design, weapon science, weaponization, nuclear material production, weapons production and military utilization, military reactors, and safeguards and security.
8. National Academy of Sciences, A Review of the Department of Energy Classification Policy and Practice (1995).
9. The rule also, in some respects, serves to strengthen the classification system. The rule confines classification to five categories of information:
10. Memorandum from Mary Ann Wallace to Anton A. Sinisgalli, February 23, 1995.
11. The National Defense Authorization Act for Fiscal Year 1996 (P.L. 104–106, February 10, 1996) enacted the “Information Technology Management Reform Act of 1996,” which will be codified at 40 U.S.C. sections 1401–1503.
12. The historical documents under the control of the DOE History Division represent a particularly good example of the value of well organized and complete finding aids. Much of this material has finding aids that were developed by the history staff in the 1970s. Because of the existence of these finding aids, historians are able to identify with precision their priorities for classification reviews. Of course, the availability of good finding aids does not by itself solve the problem of accessibility, and historians are still facing obstacles in gaining access to many of the documents because they have not yet been reviewed for declassification.
13. Paper presented by K. Marciante, DOE Oak Ridge Operations, at the
Department of Energy Records Management Conference, Las Vegas,
Nevada, June 1997.
14. One such system is the Human Radiation Experiments Information
Management System (HREX), commissioned by the Department of
Energy and developed by Argonne National Laboratory. HREX was
designed to allow users to search for imaged historical radiation documents
that have gone through the OCR (optical character recognition)
process. Additional information on HREX can be obtained at the HREX
Web site, http://hrex.dis.anl.gov. For technical questions, contact
Argonne National Laboratory via e-mail at [email protected].
15. “Record-Destruction Order Assailed; Advocacy Groups Seeking to
Overturn Rule Sue National Archivist,” The Washington Post, June 28,
1997, page A8.
16. DOE Office of Declassification, issue paper prepared in response to
specific questions posed by the Priorities Subgroup of the Openness
Advisory Panel (October 7, 1996), p. 2.
14. One such system is the Human Radiation Experiments Information Management System (HREX), commissioned by the Department of Energy and developed by Argonne National Laboratory. HREX was designed to allow users to search for imaged historical radiation documents that have gone through the OCR (optical character recognition) process. Additional information on HREX can be obtained at the HREX Web site, http://hrex.dis.anl.gov. For technical questions, contact Argonne National Laboratory via e-mail at [email protected].
15. “Record-Destruction Order Assailed; Advocacy Groups Seeking to Overturn Rule Sue National Archivist,” The Washington Post, June 28, 1997, page A8.
16. DOE Office of Declassification, issue paper prepared in response to specific questions posed by the Priorities Subgroup of the Openness Advisory Panel (October 7, 1996), p. 2.