Adopted by the Executive Committee on the advice of Council (with one abstention)
May 3, 1997
Few issues in recent time have presented the international archival community greater challenge than the proper retention, access, and preservation of records created or stored in electronic form. The Society of American Archivists (SAA), the oldest and largest association of archivists in the United States, representing more than 3,300 individuals and 500 institutions, has over the past two decades been actively engaged with the problems and opportunities electronic records present. Through its committees, task forces, commissioned publications, curricular offerings, strategic plan, and policy statements, SAA has sought to improve the ability of the archival and records management community to deal with electronic records.
One of the more controversial mechanisms used to expand the discussion of archival records issues has been litigation, most notably in Armb, et al., plaintiffs v. Executive Office of the President, et al., defendants, popularly known as the PROFS litigation. More recently, SAA was asked to join in a new lawsuit, Public Citizen et al., plaintiffs, v. John Carlin, et al., defendants (inappropriately known as PROFS 2). The suit challenges the promulgation by the Archivist of the United States on 28 August 1995 of General Records Schedule 20: Disposition of Electronic Records (GRS 20).
While SAA has endorsed in principle the joining of litigation when it advances archival interests, SAA's Governing Council at its meeting on 26 January 1997 declined the invitation to join this suit, believing that both the object of the litigation, GRS 20, and the reasoning of the legal arguments are inadequate.
Regardless of SAA's participation, the litigation may have far-reaching effects on the archival profession. Decisions made by non-archivists as a result of the suit may have a lasting impact on the nature of the historical record in the information age and on the ability of present and future researchers to use electronic records as reliable and authentic evidence of past events, facts, and actions. Therefore, both to inform the discussion surrounding this particular case, and to help advance in general our ability to meet the challenges presented by the widespread use of electronic information systems, SAA makes the following observations on the issues in the case.
1. General Records Schedule 20 is a complex document, seemingly in conflict with the current state of archival theory and practice. In addition, because of its complexity, it is likely that it will be difficult to implement (and hence be of little use to the agencies it is intended to assist). Three areas in particular stand out:
As its title suggests, GRS 20 focuses on records created and stored on a particular medium. Good record keeping ensures that adequate documentation for legal, operational, accountability, and historical purposes is created and retained, regardless of the specific technology used to create or store the records. We believe that NARA should resist definitions of "records" according to the specific technology used to create them, and should instead rely on general principles applicable to all records, regardless of format. NARA should allude to the specific characteristics of the technology that created records only when the technology raises concerns that cannot be addressed through the general principles.
Because GRS 20 lacks a consistent, clear focus, it is in conflict with the purpose of a general records schedule. As NARA's own Disposition of Federal Records: A Records Management Handbook notes, general records schedules are intended to provide disposition authority for administrative or housekeeping functions common to several or all Federal agencies. Some items in GRS 20 (items 1, 2, 4, 5, 6, 7, 8, 10, 11 and 12) relate disposition to the role of the records in the life cycle of an information system; other items (items 13, 14, and 15) base disposition on the relationship of records to copies in a record keeping system; other items (items 3 and 9) refer only to the electronic medium on which the record is stored; and some of these items may reflect programmatic activity inappropriate for inclusion in a general records schedule. This lack of clear focus makes GRS 20 different in many ways from other general records schedules issued by NARA, and undermines the effectiveness of GRS 20 as a records management tool.
Because GRS 20 is difficult for agencies to implement properly, it can be easily misused, thus defeating in part one of the functions of a general records schedule: simplifying record-keeping decisions for agencies. The most successful schedules provide agencies with specific advice on retention, and in an easily implemented fashion. Specific advice also ensures that only records truly not needed to adequately document the activity of the agency will be destroyed. GRS 20 offers instead broad and inconsistent categories that agency personnel will find hard to understand and apply. Further, the authorizations GRS 20 provides are qualified by exceptions. While the exceptions may be appropriate, implementation of some important disposition authorizations requires a very close and careful reading of the document. In sum, it requires too much intervention and decision-making by agency personnel, requiring them to understand both the definition of a record and to identify the appropriate retention period on what appears to be a case-by-case basis for individual e-mail messages. This degree of end user involvement is labor-intensive and time-consuming, and is likely to lead to inconsistent decisions and incomplete documentation.
2. The complaint, Public Citizen v. Carlin, is a seriously flawed document. It reflects a basic misunderstanding of fundamental archival principles and practice. Two areas in particular stand out:
The complaint argues that to remain useful, information must remain in the form in which it was created. While archivists are committed to facilitating access to records for research purposes, they also must maintain the reliability, authenticity, and context of the records. In many cases, archivists can maintain records in their original form, but in other cases, for programmatic, financial, or preservation needs, it may become necessary to transfer records to new media. Paper records are occasionally duplicated onto microfilm and then the originals are destroyed; important video tapes are transferred to motion picture film in order to preserve their information even though film is in many ways more difficult to manage; and electronic statistical data is routinely copied onto new media or in some cases imported into new applications. Archivists have developed procedures to ensure that when information is transferred to another medium, important contextual information about the creation and use of the records needed to support the authenticity, accuracy, and reliability of records is also preserved. The question of the format in which to save records is a difficult one and should be made only after the decision on whether to retain or destroy the record has already been made.
General records schedules, when used to manage records united by functional similarities of an administrative or housekeeping nature, ease of identification, and limited retention period, are one of the most efficient and widely-used tools available for the management of modern records. The complaint suggests that the whole concept of general records schedules should be abandoned because they do not allow for input from outsiders and because of the ease by which they can be abused by agencies. It is the belief of SAA that carefully- and precisely-defined general records schedules, limited to administrative records common to many agencies, can greatly assist in the scheduling of records, leaving the archivist and the records manager time to address more significant records series.
All archivists are struggling to establish effective programs to deal with the records generated by revolutionary new means of recording and communicating information. SAA encourages NARA to rescind its GRS 20 for the present, and work to integrate disposition instructions involving electronic records into the existing schedules that focus on government functions. In particular, NARA should first test its e-mail approaches in targeted agencies on an experimental basis to ensure that its guidance does not lead to the unwanted destruction of part of the permanent historical record. In addition, NARA should actively pursue the development of electronic record keeping systems that could automatically and effectively identify and store records created in electronic form of enduring value. SAA also encourages Public Citizen and the other plaintiffs in the case actively to seek resolution of their grievances in a venue other than the courts. NARA will not be able to address effectively the important issues in electronic records management if it must defend itself against ill-informed attacks in the court. Only by working together will we be able to develop the standards that will serve the needs of agencies, citizens, and history.