Operations Security Impact OnFebruary 13, 1998
Declassification Management Within The
Department of Defense
Booz Allen & Hamilton Inc.
900 Elkridge Landing Road
Linthicum, MD 21090
Conduct analytical studies to determine any Operations Security (OPSEC) concerns associated with declassification activities mandated by Executive Order 12958.
Executive Order 12958 effectively changes the way federal agencies manage classified national security information. It is intended to reduce the amount of information that is classified; places limits on the length of time it remains classified and facilitates the declassification of information. It sets in motion a five year time limit (April 1995 - April 2000) within which all classified information more than 25 years old and judged to be of permanent historical value shall be reviewed for declassification and declassified unless it meets certain definitive exemption criteria. All material not meeting the exemption criteria will be automatically declassified—whether or not the records have been reviewed.
To fulfill the task objective the study was to be conducted in the following parts:
Implicit in this methodology was information gathering on records management activities within DoD, the actual accession process used by NARA, methods used in agency and NARA declassification, sampling techniques, release dates, etc. A decision was made to first determine if there was an OPSEC problem with the materials prior to expending effort on a detailed review of what could have been a non-problem. Although this initially hampered a clear understanding of the procedures in use, it in fact added more validity to the data collection. The researcher had no special edge or inside knowledge at NARA, and could thus recreate data acquisition activities as experienced by any researcher with limited or no NARA experience. Any understanding of the system would initially be gleaned from publicly available educational materials, general Internet searches and whatever help a researcher could obtain in routine course from NARA personnel.
- Study DoD records which had been accessioned by the National Archives and Records Administration (NARA), declassified and made available to the public.
- Determine what, if any OPSEC concerns the release has caused. For instance, is the release of multiple documents, in one or more than one records group, from multiple agencies, a concern to national security when the review of the compilation of information reveals data that warrants additional protection.
- If there were OPSEC concerns found, evaluate the issues in view of current declassification guidelines, and make recommendations for improving the overall process.
Considering the vast amount of information available and the finite number of hours for the task. a structured work plan was constructed. The plan design ensured both completeness and flexibility to deal with issues as they surfaced. Elements of the plan were:
- Background data acquisition on declassification issues. (Government documents, client discussion, Internet, library research)
- Original research at the National Archives in DoD records already available to the public.
- Attendance at Secretary of Energy Advisory Board Meeting.
- Discussions with NARA archival personnel during research.
- Interviews with declassification managers at NARA, Defense Intelligence Agency and the U.S. Army.
The underlying principle used in approaching the review of already released material was the application of OPSEC precepts to the problem. Operations security takes a systematic look from the adversary's viewpoint. For purposes of this study the threat chosen was information acquisition to assist proliferators and terrorists in the development/use of weapons of mass destruction (Nuclear, Chemical and Biological). This targeting was selected based on items of interest delineated in EO 12958 and DoD Directive 5200.30.
In line with imitating an adversarial viewpoint was the need to maintain an openness to consider information and technology deemed obsolete by U.S. standards. The thought process was that more recent proliferators of weapons of mass destruction are normally third world nations whose scientific, logistical and management support systems are generations behind U.S. systems. However, by pursuing discarded technologies they inherit something that may be technically crude, but is proven to work. In this method, redacted scientific data, test data, organizational and management information and implementing technology based on fundamental science can significantly decrease the cost, time and skill necessary to develop the desired capabilities to successfully obtain the weapons of mass destruction capability.
Each agency of the government is assigned a number by NARA for identification purposes. This number is known as a Record Group. A researcher would therefore determine the Record Group of the agency and go about finding all available information on the identity and location of the records of interest in that Record Group. The key to record retrieval is obtaining available finding aids. The main NARA finding aid is a computer printout which lists the Record Group. the title of the file and the physical location within the archives building. This finding aid is extremely general in nature and one entry could cover hundreds of boxes of records with no clue as to the contents other than "decimal file". (See enclosures #l, 2, 3 and 4).
Approximately 37 cubic feet of records were examined during the review portion of the study. Utilizing the NARA estimate of 2500 pages per cubic foot, over 80,000 individual pages were incorporated in the review. Documents within the following Record Groups were examined. To give an indication of the variety of the examined records, the title of the series as listed in the NARA Master Location Roster and a summary of the types of documents from each are provided.
[Reviewed records are denoted with an "X" on the enclosure]. In some instances, more specific finding aids may be available for a particular group. However, there is no reliable way to predict the quality of the NARA finding aids before actually starting the physical search of the documents.
- RG 330 - Office of the Secretary of Defense
=> MLR 160A: Secret Files, 1956 (Security Classified Administrative Correspondence)
=> MLR 207A: Reports on Special Investigations conducted by Departments of the Army, Navy and Air Force, 1948-51.
=> MLR 341A: Records Concerning Organization, Budget and the Allocation of Research Development, 1946-1953.
=> MLR 342: Transcripts of Meetings 1946-51. (Effects of Scientific Research on National Security).
=> Summary: Files on supply and logistics; mobilization and production requirements for U.S. forces (ships, planes, ammunition); consumption rates; cargo handling data; budget estimates; contractor expansion requests; agendas of meetings; verbatim transcripts of advisory board meetings.
- RG 373 - Defense Intelligence Agency
=> MLR 1B: Edgewood Arsenal Technical Files Relating to Foreign Chemical. Radiological and Biological Warfare, Retired to the Defense Intelligence Agency for Reference Purposes.
=> Summary: Reports on basic scientific research; reports on conventional weapons- Japanese weapons and explosives; translations of foreign research in basic science from Sweden, Norway, Lithuania; chemical analysis of blood in gassed animals; technical analysis of small weapons, guns, gas masks and other W.W.II era material.
- RG 374 - Defense Nuclear Agency
=> MLR 3: Records of Joint Task Force Two, 1965-1970.
=> MLR 22: Inspection and Investigations File (1947-1951)- Armed Forces Special Weapons Project.
=> MLR 32A: Armed Forces Special Weapons Project Logs and Journal, 1947-1954.
=> MLR 34A: Armed Forces Special Weapons Project Decimal Files, 1946-52.
=> Summary: IG files (non-classified) - cash counts; assignment and discrimination complaints; personnel administrative matters; medical treatment bills; personal service contracting data; supply inspections; reports of survey; pay vouchers; funding estimates; testing data on Loran and radars; cost data; wiring; annual contractor reports; computer time estimates: mathematical modeling of target acquisition tests for low altitude aircraft penetrations.
- RG 457 - National Security Agency
=> MLR 9023: U.S. Army Records Relating to Cryptology, 1927-1985.
=> Summary: Analysis and summary of reports on W.W.II enemy estimates of allied troop strengths; general articles on cryptography; technical manuals and administrative notes.
The sheer amount of available text material within NARA is mind boggling. NARA estimates that it currently maintains 21 million cubic feet of material in its facilities. The Moynihan Report on Secrecy estimates that all of the Defense Department [including the component services] have one billion pages of material over 25 years old and subject to Executive Order 12958. This estimate does not include pictures, recordings or electronic data.
Just as the Internet has entwined itself in our society and altered the flow of information sharing in ways that are still becoming clear, it has also played an equally important role in the government secrecy/declassification arena. This has manifested itself in the following areas:
- Availability of information. Advances in scanning technology, digital processing and optical character recognition (OCR) assures the world wide, rapid dissemination of information. As fast as material is declassified, it is put on the Internet and is available to all. The era of information sitting in some archive available only to some scholarly researcher digging through a mountain of paper is in the past. Dissemination is now free and rapid.
- Multiplicity of copies or versions of information. The number of sites devoted to government secrecy, sensitive government programs and declassification issues is enormous and ever expanding. The sites are hosted by U.S. government agencies, universities, academic research organizations, individuals, groups and commercial entities. Enclosure #5 is a NARA finding aid available on the Internet. Enclosure #6 is a listing of declassified satellite photographs, including mission summaries and positioning data. Enclosure #7 is a Chemical Warfare (CW) website, replete with chemical compound data for CW agents. Enclosure #8 is a listing of hotlinks to Nuclear, Chemical and Biological (NBC) sites generated by a web search for declassified U.S. defense information. Enclosure #9 is the Web page for an information broker selling a CD-ROM Index and associated Abstracts of declassified defense documents. Sophisticated search engines available to anyone with Internet access, guarantee a nearly endless supply of relevant linked sites.
- Cost Savings. The ability to conduct significant research without leaving your home or place of business is a significant capability for the individuals or organizations who are performing research activities. The cost savings are not just monetary, although those are significant, but also the saving of time and manpower that was previously required to compile the data. Essentially the Internet is now allowing individuals to conduct relatively sophisticated open source analyses in relatively short periods of time.
Finding 1Information contained in four declassified documents raises OPSEC concerns. The four documents are listed below.
All of these documents were reviewed by DOE, DoD and/or NARA reviewers. In these instances, the reviewers took too narrow a view of the declassification parameters. Of the four documents, only Enclosure #12 had any material redacted. In most of the redacted material, that deletion appeared to be one figure of tonnage or hull pressure failure rates. The various reviewers do not appear to have considered the impact of the entire document as opposed to the individual pieces of information within the text. Looking at these documents from an OPSEC or adversarial perspective shows that when taken in the aggregate, these documents provide a veritable "how to" organizational file for the establishment of a nuclear weapons test program.
- Weapons Effects Test Program for Operation UPSHOT and KNOTHOLE [Enclosure #10]
- Proof Testing of A. W. Ship Countermeasures in Operation CASTLE [Enclosure #11]
- Pelican Report [Enclosure #12]
- Proposed Project Summary: Operation CASTLE [Enclosure #13].
A major problem in the development of weapons of mass destruction is procuring the organizational acumen necessary to achieve successful weaponization. Having the actual knowledge of the actual basic formulae is, in most cases, commonly known. One measurement of this is the incredibly complex logistical, organizational and test infrastructure and the necessarily long lead- and planning-times required for the development of weapons of mass destruction testing programs. Protecting the information which provides a basis for these test programs development may keep a potential proliferator at bay, or at least delay the time-line for the program development. This delay allows additional protective measures to be considered for implementation.
In addition to the organizational acumen, these documents lay out specific tests, materials and ship modifications, measurement points, survey measurements and proposed experimental procedures in one tidy package. Enclosure #12 contains a significant amount of weapon specific basic research in one place. Furthermore it names 64 additional classified and unclassified articles/books of relevant research. Not only is all the appropriate research and formulae here but there is a veritable shopping list of other reports to provide both targeting of other Record Groups and for additional open research. Although basic science should not be classified, a compilation of weapons specific material must be evaluated in a different manner. Complicating the fact is the tremendous amount of computational power available now, was not available when Enclosure 12 was written. There appears to be a high probability that a good deal of simulation testing could be accomplished based upon the research and formulae provided.
All of the documents make extensive reference to the code names of other atomic test shots and programs. Simply focusing on these code names, especially in other Record Groups such as Atomic Energy Commission, Defense Nuclear Agency, U.S. Navy, U.S. Air Force, U.S. Army, etc. could lead to other reports or perhaps even an unredacted copy or draft input within these other Record Groups. This focus could greatly reduce search parameters and yield viable results in less time. In point of fact, these documents were found in RG 330 (OSD) as a result of just a previous name reference in the files of RG 374 (DNA). Although the reference in the DNA Record Group were of no immediate intelligence value, an OPSEC approach (looking at things in the aggregate from the adversarial view) dictated the use of the code name as a search parameter for other Record Groups. Nothing more sophisticated than a pencil list and some time to conduct the search yielded the result.
Finding 2The level of organizational effort being applied by OSD and component Defense organizations to the implementation of EO 12958 is encouraging. However, at this stage of the effort it appears that it will be a very difficult task to accomplish the EO directives within the tightly mandated time-frames.
Considering the mountain of backlogged data, the string of various Executive Orders, classification/declassification policies and the relatively primitive state of records management in this backlog, the progress has been remarkable. From a zero ramp-up in late CY 1995, the various programs appear to be rapidly closing on viable methods for eventually accomplishing the task. However, with the amount of time available, it appears, there is a monumental challenge in front of the various organizations to even meet the spirit of the EO, much less meet the letter of the law.
On-site visits with Army and DIA, coupled with background data on Air Force, reveal an excellent start for the development of lateral relationships, utilization of computer and OCR technology, and innovative use of a mixture of civilian, military, contractor and reservist personnel. Each interviewee (identified at Appendix A) were appreciative of the OSD approach to concentrate on the policy level and to allow the separate agencies to solve the implementation problem from their level.
This effort provided a snapshot in time of one aspect of what DoD declassification personnel are facing. Numerous government panels and working groups are attempting to deal with the complex issues of implementation. It was not the purpose of this study to replicate any of those issues. Indeed, the system appears to have adapted within DoD and is moving towards mission completion, albeit not at a pace that will meet the goals of EO 12958.
But there has been a cost. Significant resources, both dollars and manpower, have been pumped into the program. These resources have necessarily been diverted from other activities. The promising news is that the system should begin to self-correct in three to five years as the backlog is whittled away and the new life cycle classification systems begin to take effect. By then a much reduced (resource wise) system should be doing a consistent flow of work. The following observations are provided:
1. One issue within the on-going debate is the public openness versus secrecy problem. The troubling aspect is the tendency among some "stakeholders" and non-defense or multi-missioned agencies, to address the issue as one of openness only. That should not be an option for DoD. The issue is legitimate secrecy and declassification of only that which no longer needs to be classified. U.S. citizens are ill served by any other DoD priority. Declassification decisions primarily need to be assessed in relationship of value to the adversarial organizations vice public disclosure for the sake of openness. OSD would be well served by making that the basis of DoD declassification management.Recommendations
2. The effort to provide all the declassification specialists with appropriate information is an ideal opportunity to incorporate an OPSEC review into the fundamental base of DoD declassification guidance. Looking at declassification from the Adversary's perspective would provide the declassification specialist with the methodology to review the totality of the information. This OPSEC review would allow for the continued protection of information which provides critical information to the adversary through the compilation of various small pieces of information that might not seem valuable by themselves. Currently, the declassification specialists are given the classification and declassification guidance that is available and tend to look at this information as very finite pieces of data. The potential value to the adversary needs to become the cornerstone of the guidance.
3. Similarly, the utilization of specific declassification methodologies (redaction and Pass/Fail review) need to be assessed. As noted in the previous finding, narrow point redaction can easily miss the real intelligence value of the totality of a document. Although redaction has a definite niche in very specific cases - i.e. narrow FOIA requests, it can lose its value by failing to consider the aggregate vice the pieces. Using a Pass/Fail declassification methodology keeps the totality of the information more in focus and has the added benefit of being much more cost effective to implement.
4. DoDD 5200.30 is dated. It is now 15 years old and given the new EO this may be an ideal time to revise and restructure. A key component of declassification is the matter of agency equities and how they are handled. As the components begin to consolidate, update and exchange their various program/activity declassification guides it becomes imperative that the guides be specific and readily available to all personnel, both inside and outside of DoD. It is especially critical for NARA and all the declassification specialists to have the latest DoD and component guidance.
5. DoD needs to study and assess the use of the Internet in the overall departmental declassification strategy. As noted earlier, the Internet has become an integral part of the entire secrecy/declassification issue. The question becomes how to effectively utilize this tool to advance DoD declassification goals. First and foremost DoD must identify a clear set of specific goals; assess the strengths and weaknesses of the Internet and then devise a strategy to reach those goals. For the sake of a hypothetical model assume that the DoD goals are:
=> Conduct a declassification review of all 25 year old materialBased on these goals one would then assess the strengths and weaknesses of the Internet.
=> Identify and segregate that material which should remain classified
=> Declassify and make available to the general public that material which no longer needs protection.
=> Manage FOIA reviews in the most cost effective manner
=> The use of the Internet provides a rapid and cost effective method for the dissemination of unclassified information.Weaknesses:
=> The use of the Internet could reduce the unrestrained public appetite for "secrets" by providing good faith distraction material
=> The use of the Internet could channel public interest towards already appropriately declassified material and possibly lessen FOIA requests.
=> The use of the Internet could have rebound effect and fuel a more voracious public demand for ever more material. May facilitate more FOIA requests by providing a shopping list of available materials.A strategy could then be devised by DoD and the components, based upon this evaluation, to implement a coherent and complimentary plan to achieve the declassification goals. For example:
=> The use of the Internet could overwhelm the administrative system that processes inquiries. By providing documents that have been recently reviewed and declassified, it can magnify imperfections in the declassification system by making available declassified material out of historical context.
=> Information that is available over the Internet magnifies the imperfections of the declassification system. For instance if a document should appear on a website in full text, but is later shown as a redacted document or is refused to be released for classification reasons from another source, it would bring the entire system into question.
=> Openness: Discussion of balance between necessary secrecy and openness- i.e., continued classification of old nuclear test data to keep out of terrorist hands.Central direction by OSD and close coordination with complimentary action by the components could effectively enhance the accomplishment of DoD declassification goals.
=> Diversion: List of interesting declassified material - i.e. Kennedy assassination data.
=> Cost effectiveness: Provide referrals within the website to available holdings at NARA. This would provide the researcher with valuable information and with a research direction that might prove very fruitful for them and potentially satisfy their requirement without taxing other parts of the system.
6. Classified RG330 records at NARA should be reviewed by OSD. Although not part of this study, a survey of the group, followed by a risk management approach to a declassification review should be conducted. Preferably, this should be done under OSD auspices to insure that the correct emphasis and thoroughness is applied.
7. Finally, this study seems to confirm that a great deal of material can certainly be declassified within 25 years. Conversely, a small but significant amount may have to remain classified indefinitely. However, periodic reviews driven by viable threat and risk management techniques should be the only determinate for that decision. As shown above, some information, especially in the weapons of mass destruction category, does not lose its value over time. Indeed some of the least sophisticated and oldest material is precisely of more value to a potential proliferator or terrorist. It only has to work. The simpler the better, especially if your resources, scientific and logistical infrastructure is lacking.
1. Update DoD declassification guidance. The inclusion of current general declassification instructions that incorporate an OPSEC process and promotes the application of a Pass/Fail declassification system, coupled with continued emphasis on the update and dissemination of specific component and project declassification guidelines ensures an effective declassification program.
2. Develop a comprehensive plan to utilize the Internet to achieve DoD declassification goals. This plan should be developed in concert with the components to ensure a unified approach.
3. Provide current DoD direction for the declassification of specific subject areas. The directive DoDD 5200.30, provides basic guidelines to the declassification community. Currently the directive is being used to provide a first cut at what subject areas require further review by specific agency/organization declassification specialists. The inclusion of current general instructions into a restructured directive or into the DoD Declassification Plan and the continued emphasis on the various agencies/organizations to ensure current declassification guidance is available for the specific projects and programs is essential for effective declassification of material.
4. Conduct a review of the declassified and classified RG330 records at NARA. Classified records would be reviewed under the EO year 2000 provisions. It is recommended that a complete review of the records be accomplished. However, recognizing resource limitations within the declassification community, a statistically viable sampling of the records, with the OPSEC concerns being addressed throughout the review, would provide information on what currently sensitive information has been inadvertently released to the general public. This knowledge would provide a means to provide the declassification specialists with more specific guidance on how to review material in the future and would allow specific items of information to brought back under control to minimize the damage that may have occurred.
Declassification of information that no longer needs to be protected is an excellent objective. Indeed a very large amount of the material reviewed appeared to be of no use - intelligence or otherwise. The 85 to 95% declassification ratio mentioned in my background reading seemed to hold up. However that is to be expected in records that are 45 to 50 years old. This does not mean releasing information solely based on age or extremely narrow review criteria. Some information or collection of information, especially in the area of nuclear weapons, does not lose its value with the passage of time. Indeed some of the least sophisticated and oldest material may be of most value to a potential proliferator. For instance, according to the DOE, Secretary of Energy Advisory Board (1997), the Iraqi nuclear program was based on obsolete U.S. 1940's technology.
Given all the above we need to keep certain facts in mind.
- Some material, no matter how old, deserves to remain protected.
- As we come up to the timeline of material to review, the 85 to 95% declassification ratio will probably drop. This is simply a matter of relevancy to the present.
- An OPSEC oriented approach to declassification management will markedly enhance the program. The declassification specialists should review the documents from the view of the adversary and consider what information the adversary needs to effectively advance their objectives.
Appendix AJean Schauble, Director, Records Declassification Division, National Archives And Records Administration, College Park, MD.
LTC Gary Moore, Operations Officer, U.S. Army Declassification Activity, Arlington, VA.
[name of one interviewee from the Defense Intelligence Agency DELETED]