The Honorable Martha B. Gould
Chairperson
National Commission on Libraries and Information Science
1110 Vermont Ave. NW, Suite 820
Washington, DC 20005-3552

Dear Ms. Gould:

The National Archives and Records Administration (NARA) offers the following comments on the November 27, 2000, draft of the National Commission on Libraries and Information Science (NCLIS) report, Comprehensive Assessment of Public Information Dissemination. As we have stated in the past, NARA strongly supports improvement of public access to public government information; the continuation of the National Technical Information Service (NTIS) functions in the Executive Branch; and appropriations for the NTIS “public good” functions specified in the draft report. We also agree with the NCLIS view that public information is a strategic national resource. However, overall, we found the draft NCLIS proposal to be seriously flawed and we do not endorse the proposal to establish the Public Information Resources Agency (PIRA) as a new independent agency in the Executive Branch.

We recognize that NCLIS set itself a very ambitious goal to conduct a comprehensive study of Federal dissemination of public information in an extremely tight time frame. Unfortunately, the report reflects sweeping generalizations, often based on anecdotes or selective quotes instead of the detailed analysis that we hoped for when we commented on the Executive Summary. The conclusion that a new central information services agency must be created to cure current problems with the Federal government’s implementation of Title 44 requirements is not well supported.

We are very concerned that the report shows a fundamental lack of understanding of NARA’s mission and statutory authorities. There also is little recognition that information products start out as records created by Federal agencies to facilitate the conduct of business, and therefore must be managed as records, regardless of how many copies may be disseminated under other statutory authorities. We have particular problems with recommendations 24 and 29 as they pertain to agency and NARA responsibilities.

Recommendation 24 shows a lack of understanding of how the records schedule process works and its relation to records disposition. Whom does NCLIS believe will determine the metadata elements? Under the Federal Records Act, only NARA can determine such metadata, in consultation with interested parties. The report also needs to indicate how this information is to be shared so that it will “aid in the searching and acquisition of government information, preferably on the Internet.”

Recommendation 29 is unacceptable to NARA. As NCLIS details its vision in this recommendation, it proposes to fundamentally gut the mission, mandates, and responsibilities of NARA, and the authorities of the Archivist of the United States under the Federal Records Act. We disagree strongly with the proposal that PIRA would assume agencies’ responsibility to transfer permanent records to NARA, and the implication that PIRA would decide the appropriate time and format as well as whether to transfer the records to NARA.

The imprecision in using the terms “public information resources,” “data,” “information products,” and “records” throughout the report narrative makes interpretation of this recommendation more difficult. If “public information resources” used in this recommendation is interpreted broadly, as NCLIS does earlier in the report, all permanent records may be covered by the proposal, not just government publications and reports.

The report is so vaguely worded that “public information resources” could include copies of very transitory records, such as lists of agency credit card holders posted on an agency’s EFOIA reading room site. If NCLIS intends that PIRA would retain such records for permanent public access, this would result in an enormous unnecessary expense.

We were troubled by the report’s promotion of the term “archiving” to mean something other than preservation of permanent records in the National Archives of the United States (see especially p. 8). We believe that this muddling of the meaning of the term only compounds the current confusion over the distinct and separate agency responsibilities to provide copies of information dissemination products to the FDLP system and to transfer permanent records to NARA.

We also have a serious concern with the proposed building of an Information Life Cycle Management software module discussed on page 50 of the report, finding 6.D on page 75, and in recommendation 18. The report ignores significant work being done in this area, including the DOD 5015.2 standard and NARA’s Electronic Records Archives (ERA) project. As we stated in our November 14, 2000, comments on the White Paper, we believe that the proposal needs significant further work to be a feasible NCLIS recommendation.

We are very concerned that the critical recommendations relating to the continued operation of NTIS in the Executive Branch may not be implemented while various stakeholders within and outside the Government debate and refine your proposals for PIRA and separate centralized information resources agencies in the Legislative and Judicial branches. We note that if NTIS does go out of business, its permanent records, including Federal publications it holds, must – by law – be deposited with NARA, not with the Library of Congress. In fact, NARA and NTIS have worked this out.

We also offer the following comments on statements relating to NARA:

• We would not include NARA Bulletin 2000-02 (p. 14) in the list of “landmark government initiatives,” and the description of the bulletin is inaccurate.
• The report is in error on page 59 in stating that NARA only accepts 62,050 bpi tape as a transfer medium. NARA has long received permanent records via CD-ROM and is preparing to accept electronic file transmission. Unfortunately, the URLs for Appendix 26 and the agency survey results were incorrect, so we were unable to determine the source of the misinformation.
• Mention of the Federal Register as a vehicle for government information dissemination and as a “success story” for interagency sharing of government information for the public good (page 60) is notably absent in the report. Federal Register publications account for almost 90 per cent of the traffic on GPO Access.
• We disagree with finding 5.0 on page 66. If, as the report asserts, “[m]any agencies do not schedule all of their publications and important documents as permanent records,” then the agencies are not scheduling their records or are applying the schedules incorrectly. As a class, publications are almost always regarded as records, and NARA guidance on scheduling records uses publications as a classic example of records worthy of permanent retention. Other “important documents” would also be appraised for possible permanent retention. By what process has NCLIS determined that publications and important documents are not scheduled as permanent?
• For the same reasons, we do not accept the statement on page 74 that “The National Archives and Records Administration (NARA) has responsibility for the retention and preservation of the records of government, but not necessarily for all publications of the federal government.”
• We disagree with a number of statements in the paragraph on page 74 that begins “Let us be a bit more specific…”:

First, the authors do not appear to understand that depository libraries and archives function differently because they have different missions and principles under which they operate. Because of these differences, there will necessarily be “inconsistencies” between the FDLP and NARA operations, as they serve the public and agencies in different ways to meet their different missions. The gratuitous note that NARA reproduction service is “not usually on a timely basis, however” is both incorrect and unnecessary to the point being made.

NARA does not now accession records at the beginning or in the middle of the government information life cycle. It is entirely likely that this will change with the advent of ERA. Moreover, we collect information at all stages of the life cycle through the scheduling and appraisal process. We also have long called on agencies to implement information management controls during the conceptual stage of electronic systems development.

Finally, NAIL is a prototype of our Archival Research Catalog (ARC), currently under development. Neither NAIL nor ARC is intended to be an item level catalog of all of the individual files held in the National Archives of the United States. We agree, however, that such bibliographic metadata is needed to manage information dissemination products. NARA is supportive of approaches that would incorporate such metadata into electronic records management systems.

We hope you find these comments helpful. Because we do not support creation of the proposed PIRA, we are not providing detailed comments on the proposed draft legislation.

Sincerely,

JOHN W. CARLIN
Archivist of the United States