Best Practices for Document Management in an Emerging Digital Environment
Richard E. Barry, Barry Associates
[This paper was originally published in the Records Management Bulletin, the journal of the Great Britain Records Management Society (RMS),Issue No. 63, August 1994, p.11, and is republished here with the kind permission of the publisher.]
British Common Law did not require or allow the use of written forms as evidence until sometime after the Middle Ages. After all, documents couldn't talk. Nor could they take oaths. People who could talk, and talk under oath, were seen to offer the most trustworthy evidence. That all changed, of course, with the dramatic increases in commerce, changes in the use of documentation to reflect business transactions following the Middle Ages, printing methods and corresponding changes in literacy. Initially, records were regarded as "hearsay" evidence or statements from sources other than live people. Only in more modern times have exceptions to the hearsay laws been adopted which permit records to be used in evidence under specific conditions. There are two obvious but important messages emanating from the original Common Law story that are as current and important today as they were in early times -- the importance of jurisdiction and of trustworthiness of documents, records and recordkeeping systems.
First decisions about electronic document and records, and their management systems, must be taken in the context of the legal and administrative jurisdictions under which the public or private enterprise must operate and the laws or rules governing that jurisdiction. Simply stated, this means that the management of electronic documents and records must comply with applicable international, national and local legal codes, regulations, treaties, agreements, established practices of the records and archives professions (which vary among European, North American and other traditions) and other applicable administrative rules which are not otherwise superseded by legal requirements. For example, in the U.S., the Federal Uniform Rules of Evidence provide that "any printout or other output readable by sight, shown to reflect the data accurately, is an 'original.'" Similarly, Section 3 the Swedish Archives Act of 1990 provides that "computer records which are available to several authorities, thus constituting official records, shall only form archives at one of these authorities, in the first instance the authority which is responsible for the major part of such records." Swedish National Archives orders concerning technical requirements on materials and methods for long term preservation of records have been adapted to the corresponding rules within the European Community.
For purposes of this paper, jurisdiction is meant to embrace legal or administrative governance boundaries in political or organizational terms and the form of rules imposed by political or organizational entities. The former may include one or more of the following: international, religious, regional, national, state or provincial, local (county, city, parish, town) or organizational entities. The rules they imposed may be in many forms: laws; treaties or agreements; executive or religious orders; national archives requirements; local codes and ordinances; organizational policies, directives and procedures; commonly held professional standards, etc.
The requirement to comply with established records laws and other rules is well known and understood by most records managers and archivists. It is less well known or appreciated by professionals in the information management and technology (IMT) field, because this group traditionally has not had jurisdiction over, or interest in, organizational paper records, and because the IMT tradition is one which focuses more on media than on content, whereas the archivist and records manager must worry about both. Too frequently, both professions fail to understand the most important borderline area between their professions: designing good records management and archival practices into electronic systems, in whatever form -- data, text, image, audio or video -- and in whatever medium -- paper, microform, analog or digital.
If an electronic document management system is being introduced by either the IMT or records and archives management organizations in an enterprise, it is essential that this be done with the consultation and involvement of the other, as well as key users of enterprise records assets. This is important for two reasons: it may happen that certain documents which are copies of official records will find their ways into a "vanilla" electronic document management system (EDMS) that is not intended to meet the organization's electronic records management (ERM) needs and, depending on jurisdiction, may constitute an original and have to meet the same disposition requirements as the original records; and because the common sense thing to do may be to use the EDMS as the vehicle for managing electronic records. Either way, the IMT or records professional, trying to do it alone, is walking down the slippery slope of legal requirements which may be imposed by the legal jurisdiction under which the organization is required to operate. This paper will not deal further with the specifics of legal requirements for admissibility or discovery of evidence under any particular jurisdiction. It will, however, address how jurisdiction and rules fit into the shopping basket of "best practices" for developing a trustworthy electronic records system.
Second, the early Common Law testifies to the central importance of trustworthiness in the use of information. In modern information management terms, this refers to the capture, maintenance, preservation and presentation of information. This is important from a legal perspective, because trustworthiness is an essential characteristic before information can be considered admissible in a court of law as evidence. Moreover, the human impact of the use of information in a court is potentially very far reaching in both positive and negative senses of the terms. Trustworthiness is important from an administrative perspective because of its implications for management accountability operational continuity, and historical legacy and research. Records are kept for both legal and administrative purposes. The term "information" is used, because documents (in the usual sense of the word) and records (in the records management sense of the word) constitute subsets or instances of information. This is important not just from a theoretical perspective but because, as we enter the realm of electronic documents, including records, the broader information management and technology architecture of an organization must take fully into account the organization's legal and operational needs for managing these assets. Often, they do not. Thus, the "best practices" needed to create a defensible digital document and records management program (including organization, policies, procedures, practices and human elements) and system are in the aggregate all centered around one thing: being trustworthy, being seen to be trustworthy and, together, being able to produce, or reproduce, information which can be trusted to be used in the day-to-day operations of the organization, by the courts of the land, and by future researchers and historians.
There is a tendency to focus on how to make information systems rigorous enough to withstand the scrutiny of the courts, which is what the remaining discussion of "best practices" centers upon. However, to ensure that information is trustworthy means that the entirety of the information management program is credible. That goes well beyond the characteristics of the information system(s) and not all of it is under the direct control of the IMT manager, archivist or records manager. It includes aspects of organizational culture; formal organizational arrangements for the management of documents; the clarity of responsibilities; policies, procedures and practices that are in place; and the number, quality, skills mix and proficiency of the people responsible for husbanding an organization's information assets. It is not enough to have the above areas able to demonstrate trustworthiness. In the end, the issue boils down to the quality of implementation of the whole program. The policies, organizational structures, procedures and people are all in place, and they are "walking the talk."
The following sections outline some of the more important practices which an organization can follow to help achieve trustworthiness in its electronic records and related electronic records management systems. There is no recipe for total success in demonstrating the trustworthiness of an electronic records system. Neither is there for paper records systems. In fact, it may be argued that, depending on the jurisdiction's laws, electronic records may be dealt with under current regulations without any special problems arising. Even in such cases, there may be the perception that electronic records are less trustworthy. Therefore more rigorous minimal steps may be needed to provide necessary assurances to officers of the court that reasonable and prudent steps are being taken which assure trustworthiness at levels at least as good as what they have become accustomed to in the paper world. Indeed a persuasive case has been made in a technical report of the U.S. Association for Information and Image Management that it is possible to achieve much improved trustworthiness with electronic systems than is feasible in paper-based systems. Paper records typically exist over the long term in original form only and are more easily subject to theft without timely detection, destruction by natural causes, etc. They often use paper or ink which has a relatively short life expectancy, are not part of a clearly defined process or life-cycle system, and are rarely subject to rigorous security measures or audits. By contrast, it is feasible to develop electronic systems which overcome most of these shortcomings in paper-based systems. The AIIM report concludes: "In sum, properly designed, implemented and maintained information technology systems are capable of producing records that are more reliable and accurate than paper-based systems." What is necessary, and what is the trick, is to develop the electronic records system in such a manner as to overcome paper system shortcomings through the incorporation of some best-practices design approaches. Unfortunately, as few if any fully functional electronic records systems have been implemented, it is possible only to provide our best understandings of what constitutes best practices today which might be feasible to automate or assist through computer-based systems.
Provide for Efficient Management of Documents, Rules and Applications
The architecture of the EDMS should provide for data independence in the actual Document Data Base (DDB) where the documents are stored. Stated another way, the electronic representations of documents should be separate from any Applications Programs (AP) or associated Applications Program Interfaces (APIs) which may operate on the documents or any Rules Data Base (RDB) against which documents may be tested. Similarly the AP and RDB should be separate entities. This approach permits the use of different future systems (APs and APIs) and changes in the RDB with the same DDB.
Document and Maintain Governing Laws, Regulations,
Standards and Systems for Electronic Records Management
A complete inventory of laws and regulations and other agreements which govern the definition, creation, capture, use, preservation and disposition of records should be maintained. Apart from legal admissibility, some of the most difficult and controversial issues concerning electronic records have to do with the balancing of legal, organizational and personal interests. These are epitomized in some of the principal emerging issues:
· addressing “best practices” as a systems issue
· ensuring the integrity (authenticity, security and validity) of electronic records
· privacy vs. organizational ownership and access to information
· legal requirements governing public access to records
· corporate awareness and due diligence
· intellectual property rights
· preservation of and future access to records in electronic form
· organizational location of archives and records management functions. While no general guidance can be provided to meet the needs of every public and private sector organization, a cardinal rule here is to make sure that the employees of the organization are well informed concerning these issues and the organization's policies and procedures for dealing with them.
For purposes of electronic records, governing rules should be maintained both in a human-readable document and in machine-readable form. The former represents the governance system in a way that anyone can easily read. The latter provides a basis for automated use of the rules, where appropriate, and can be achieved through various information engineering methods and tables which may be embedded in the RDB.
Consider Future Forms of Records and Recordkeeping
Think about the possible impact of emerging technologies such as electronic mail (e-mail), voice mail (v-mail), electronic forms, groupware, workflow systems, roomware, pen-based technologies, "video documents", location-independent computing, etc. Will these technologies be in response to changing work patterns or are they emerging in response to the need for better and more efficient work patterns? For example, large commercial organizations are beginning to use the "Arizona-Room" type conference room and software system to conduct business decision meetings in an attempt to make meeting results more effective and meetings more efficiently run. The software associated with these rooms maintains a record of the proceedings of decision meetings, thus obviating the necessity for traditional "memos to file" containing minutes of such meetings. E-mail began as a substitute for telephone tag. Today it is commonly used to carry on substantive business. V-mail, presently in its early days, is now commonly used to substitute for telephone tag, much as was the case in the early days of e-mail. Will vmail go the same way and become a potent medium for conveying substantive business communications? Will these changes require a different definition of the Document? Will different recordkeeping rules, practices or systems be needed? How will trustworthiness of information and future admissibility of records as legal evidence be assured with changing work patterns and technologies? The prudent information manager will think about these things today and try to assure that system design doors are not unnecessarily closed today that may need to be opened tomorrow.
Link Records to Organizational Business Processes
For records to be records, and to be admissible in most judicial systems, they must be produced and managed as part of the normal course of business. It is not acceptable, for example, for records to be destroyed according to some retention schedule which was created when an organization has been sued for negligence. The most trustworthy records are those which are produced as part of essential organizational business processes. Traditional records and archives systems maintain records by series and provenance. While automated systems may be programmed to do this, they may also be used to provide even closer links to the core business of the organization, thus further reinforcing the trustworthiness of the records assets. This can be done by linking record series to major business processes. Business processes are more stable platforms than organizations or provenance; however, in electronic systems the two approaches need not be mutually exclusive.
Minimize and Provide Discovery Mechanisms for the Occurrence of Malicious or Unintentional Alteration or Destruction of Records
Both procedural and computer-based approaches may be applied to minimize accidental or malicious alteration or destruction of records. From a policy perspective, the organization must decide who is responsible for the maintenance of electronic records. Will they be maintained in the APs which produce them as has been suggested by some David Bearman and others in the field? Or does it make more sense to provide for a central electronic records system, either logically, physically (in a separate computer under control of the archivist or records manager), or both? Or is some combination of centralized and distributed approaches more desirable? From a systems perspective, key-stroke monitoring or more rigorous computer auditing approaches can be employed to document alterations and destruction actions and authorizations. Write-Once-Read-Many (WORM) technology that prevents alteration can be used for storing records. Separate dual storage of all records (or perhaps only archival records) can both protect against loss due to natural disaster or malicious destruction, and can provide a basis for comparison of documents to establish authenticity and absence of alteration. While dual systems are expensive, they may be desirable for compelling disaster recovery reasons. Few who lost vital records in the bombed World Trade Center, or the flooded Chicago Board of Trade Building have problems with the cost-justification of dual systems.
Apply Appropriate Retention Schedules
Electronic records, like any other, must be keyed to an appropriate retention schedule, and that schedule must be consistently implemented. Both functions may be automated. This means that steps have to be taken to ensure that electronic records are migratable over time and through changes in technological environments.
Provide for Equal Treatment of All Copies of Records
In most jurisdictions, it is essential to treat all copies of the same document in the same manner when it comes to disposition management. If disposition management functions are part of the automated system -- the most efficient approach -- then all copies of the electronic documents must be destroyed when the retention schedule calls for it. How this is managed is dependent upon the system architecture. In some electronic mail systems, for example, each recipient of a document has a literal copy of the document in his or her storage space (e.g., in the distributed, PC-based, The Coordinator System). In other systems, there is only one copy of the document which is maintained in a shared space in a centralized host computer, and "copies" are virtual, i.e., simply a group of pointers associated with the document which represent users with a copy or "view" of the document (e.g., DEC's All-in-1 system). This also highlights the need to link electronic records to paper records through a file schema equivalency table or other such mechanism.
Capture Context and Content of Records
Records have both content (information) and context (meta-information about the document). Both are needed to determine the authenticity of the record and the competency of the source. Much of the latter information may be captured in the document profile which may contain information which is supplied by the creator or the system and includes such information as:
· author and date of creation
· unique identifier
· parent business process
· provenance or office of primary responsibility
· document type
· language (both linguistic and software)
· revision dates
· retention schedule or projected/actual dates of conversion to other media or storage levels (on-line/off-line) and destruction
· security classification and projected/actual dates of downgrading
The Document Profile and Document Content together constitute the Document. Having clearly spelled out and programmed systems for managing document components provides strong evidence of the trustworthiness of documents being managed in any electronic document management system -- probably much better than is possible (or affordable) in most paper-based systems. While it is possible to retrofit information systems to provide the functionality suggested above, one of the best ways to assure system and information credibility is to design new systems from the beginning to take account of electronic records and archival needs.,7
Document and Permanently Retain Records of Records Disposition Actions Taken
As with paper records systems, it is essential that the records of disposition actions taken on electronic records be themselves treated as permanent records. This can be facilitated in electronic systems. One approach is to keep the Document Profile when the Document Contents are destroyed, with an appropriate notation in the profile of the date of destruction, which can also be provided automatically. The profile of such documents can then be maintained permanently in the electronic archives as evidence that the disposition action was taken in accordance with applicable retention schedules and rules.
Provide Measures to Guard Against Common Assaults on Electronic Records Systems
One of the best insurance policy for electronic records systems is to have bulletproof methods for defending an organization's electronic recordkeeping practices against the likely attacks of litigants whose aim it is to demonstrate the untrustworthiness of a system. The AIIM Technical Report states: "Common assaults on the integrity of computer-based technologies include challenges to: a) the source of the input data and the processes for transcribing it to machine-readable form; b) the process that creates, edits and updates the files; c) the process that produces the output or retrieves the records; and d) the reliability of the equipment and vendor supplied software that systematically manages the internal system processes...Of particular importance in fending off these assaults is to assure the existence of up-to-date documentation that fully and accurately describes the procedural controls employed in producing the records. Additionally, records managers or custodians must be prepared to describe these controls in laymen's terms, showing how they count for every link involved in producing the records." Henry H. Perritt, Jr., Professor of Law at Villanova University sums it up this way:
There is no reason to be concerned that well-designed electronic records systems present any new problems with admissibility of electronic records under the rules of evidence....A major difficulty in appropriate records management strategies is a lack of familiarity by the government's lawyers with the technology, and conclusions reached about legal treatment of the technology.
Rick Barry is President of Barry Associates, Arlington, Virginia, USA. He consults and gives seminars internationally on information management, electronic document and records management. He was formerly with the World Bank as chief of the office systems where he was responsible for managing information technology standards and services and, subsequently, chief of information services where, inter alia, he was responsible for the management of World Bank records and archives. During that time (1987-90), he also served as Chairperson of the United Nations inter-agency, inter-disciplinary, Technical Panel on Electronic Records Management (TP/REM), under the auspices of the UN Advisory Committee for the Co-ordination of Information Systems (ACCIS). The results of the two year study were published by ACCIS in: Management of Electronic Records: Issues and Guidelines, 1990, U.N., N.Y and Geneva. He contributed to a follow-up task in the preparation of: Management of Electronic Records: Curriculum Materials, 1992, U.N., N.Y. and Geneva.
 Recordkeeping Requirements, by Donald S. Skupsky, JD, CRM, Information Requirements Clearinghouse, Denver, Colorado, 1991, pp. 135-138.
 Skupsky, supra note 2, p. 145.
 "The Impact of Information Technology on the Freedom of Press Act, the Data Protection Act and the Archives Act" by Claes Granstrom, National Archives, Sweden, unpublished manuscript to be published shortly in Janus, the International Council of Archivist.
 "Electronic Records Management and Archives," by Henry H. Perritt, Jr., in University of Pittsburgh Law Review, Volume 53, Number 4, Summer 1992, pp.998.
 "Guidelines for Admissibility of Records Produced by Information Technology Systems as Evidence," AIIM Technical Report TR31-1992.
 "Electronic Documents and Records Management Systems: Toward a Methodology for Requirements Definition," by Richard E. Barry, in Information Management & Technology, the journal of Cimtech and UKAIIM, Vol. 27, No. 6, November 1994, London, p. 251.
 "Getting it Right: Managing Organizations in a Runaway Electronic Age," by Richard E. Barry in Information Handling in Offices and Archives, Angelika Menne-Haritz, Ed., K.G. Saur, London, 1993, pp. 27-55.
 Author and editor of quarterly newsletter, Archives and Museum Informatics, 5501 Walnut St., Suite 203, Pittsburgh PA, 15232-2311.
 Management of Electronic Records: Issues and Guidelines, 1990, U.N., N.Y and Geneva.
 Research Issues in Electronic Records, National Historical Publications and Research Commission, Washington, D.C., 1991.
 AIIM Technical Report TR31-1992, supra note 6, pp. 5-6.
 Perritt, supra note 5,