May 24, 1996
Michael F. DiMario
U.S. Government Printing Office
732 N. Capitol Street, N.W.
Washington, D.C. 20401
Dear Mr. DiMario:
We appreciate this opportunity to offer some final comments on the Report to the Congress: Study to Identify Measures for a Successful Transition to a More Electronic Federal Depository Library Program. On behalf of the American Association of Law Libraries, the American Library Association, the Association of Research Libraries and the Special Libraries Association, we again thank you for including us as advisors to this very important study.
Our earlier letters to you on the March draft report to Congress and on The Electronic Federal Depository Library Program: Transition Plan, FY 1996-FY 1998 issued in December, have already articulated many of our concerns. These are very important issues, and include bibliographic control, long term access, preservation and authenticity, to name but a few. We firmly believe that these issues must be decisively addressed before the transition to a predominately electronic program proceeds any further.
In addition, we believe that Task 1A, the "Technical analysis by a Federally-funded research and development center (FFRDC)" was crucial to determining the most cost effective way to implement the more electronic program. We reiterate our belief that the Technical Implementation Assistance study (Executive Summary, Appendix A) must be carried out to provide necessary analytical data on technological issues including hardware, software, and communications options. The surveys of depository libraries and agencies will assist GPO in making informed decisions on how the transition can reasonably be achieved.
We have some additional comments regarding the technological infrastructure. Planning for technological change is never easy, but the enormous technological change in the underlying information infrastructure makes the job even more difficult. There are few certainties about what the infrastructure will look like next year, much less five to ten years from now. For instance, five years ago no one predicted the growth and range of use of the Internet that has occurred since that time. (We refer you to the recent report of the Computer Science and Telecommunications Board of the National Research Council, titled The Unpredictable Certainty.)
We do know that, rapid as the growth has been, the evolution of a robust widespread and high-speed national infrastructure will be slower than what seems to be anticipated in the strategic plan. Capital investments must be made in the underlying technology, marketable applications need to be developed to stimulate private investments, and users must invest in technology and training at their end, in order to take advantage of new services. This takes time.
We also know that the basic architecture of the future infrastructure is still undetermined, and may take some time to settle down. Will highly centralized services and resources be most economic and effective, or will technology favor distributed resources? Will application software reside in the net or will it be in the users' computers? How will the economics of high quality printing evolve? The answers to all of the questions may well change over time as new innovations reach the market, as new discoveries are made in the laboratories, and as users such as libraries, publishers, and government agencies find new ways to use information technology.
Thus, the following two points are critical:
First, the transition plan should be flexible and evolving and not be overly dependent on particular technological characteristics and projections. In brief, it should not put all of its eggs in one technological basket and not be wedded to rigid timetables.
Second, the transition should incorporate a formal and continuing process of technology scanning and evaluation that moves forward as the project moves forward. We have recommended all along that the GPO should be allowed to conduct such technology evaluations. Not only is it a critical need, it is an on-going one.
In addition to these comments, we are attaching responses to several of the task reports included in the study. As advisors, we rely on our membership for input on issues of such critical concern to the future of the Federal Depository Library Program. These comments have been drafted and discussed by highly-skilled members of our four associations who, as depository librarians, have first-hand knowledge of the impact of these important issues on their institutions and their users. The task force reports contain valuable comments and suggestions that we hope will be taken into consideration as we move together towards a more electronic program.
Lastly, we join you in affirming the Principles For Federal Government Information and the Mission and Goals For the Federal Depository Library Program as articulated in the draft report to Congress (Sections III and IV). The Federal Depository Library Program has proven to be a highly successful partnership for 139 years between the government and libraries located throughout our nation in almost every Congressional district. The benefits of the program contribute directly to the knowledge of citizens everywhere about the activities of their government, and to the economic well-being of our nation.
We are pleased to have participated as advisors throughout the lengthy study process. We hope that the dialogue among the various partner agencies and the depository community will continue as plans are implemented for the shift to a more electronic FDLP. Thank you very much for considering our concerns during the study process. Please do not hesitate to contact any one of us if we can be of further assistance.
Robert L. Oakley
Washington Affairs Representative
American Association of Law Libraries
Carol C. Henderson
Executive Director - Washington Office
American Library Association
Prudence S. Adler
Assistant Executive Director
Association of Research Libraries
David R. Bender
Special Libraries Association
cc: Members, House and Senate Legislative Appropriations Subcommittee
Chair and Ranking Minority Member, House and Senate Authorizing Committees
Ms. Linda Kemp, Staff Director, Joint Committee on Printing
TASK 5: Evaluation of Incentives for Publishing
Agencies to Migrate From Print Products to Electronic Format.
ABSTRACT: The full participation by publishing agencies is essential to the success of any government information dissemination program, yet it is important to recognize that agencies have many responsibilities and many pressures on limited budgets. The only positive incentive for agencies to convert depository materials to electronic formats will be a system which is as automatic and cost effective for them as the traditional program. Since provision of information to depository libraries is not a major part of agency missions, the incentives to adopt electronic publishing must come from a broader vision of the value of an informed citizenry; data which identifies current progress and barriers; and directives to agencies which make the government commitment to information access very clear. In the electronic environment there is a need for central coordination of public access to government information. The existence of a program which would provide leadership in standards, cataloging, and long-term access could in itself be an incentive for agencies to use electronic publishing as a cost-effective way of carrying out missions while assuring public access to information.
The issues raised in Task 5 are very important ones, since without full participation by publishing agencies no government information dissemination program can be completely successful. Depository libraries have tried to find ways to develop communication channels with as many agencies as possible. Since the depositories serve users of agency information who may not be recognized by the agencies as their primary users, depository librarians are in a position to communicate user needs, suggest improvements in agency products and software, and to recommend agency publications and electronic resources to potential users and buyers.
The legislative requirements for the GPO study ask for a study which "surveys current and future dissemination plans of executive branch agencies." Without the data which would have been gathered by the technical analysis of an FFRDC (Task 1), it is not possible to identify with much accuracy the progress which is being made by agencies or the barriers which might lead to the identification of incentives. The mention in the Strategic Plan that a survey will be part of the Technical Implementation Assistance is very positive and this survey will be useful in expanding on the incentives identified in Task 5. The task assumes that agencies should be migrating from print to electronic formats. Many agencies are making major strides in that direction. On the other hand, there are some publications which agencies will decide are most useful to their primary clientele in paper format. In such cases, it will be important to weigh both the costs of reproduction and distribution in paper format or the cost of electronic conversion, and the usefulness of the final product. If the agency has no need of its own to provide an electronic version of a particular publication to meet its mission, another entity such as GPO will need to absorb the costs of electronic conversion if that format is to be provided to depositories.
The Task 5 report makes a powerful point in explaining why the Federal Depository Library Program (FDLP) works so smoothly for agencies in the traditional formats. The GPO reproduces the necessary extra copies without any effort on the part of the agencies, and Congressional appropriations pay the costs of reproduction and distribution. The publications are made available across the country in a way that is simple and cost-effective. The only positive incentive for agencies to convert to electronic formats will be a system which is equally as automatic and cost-effective for them. An additional incentive would be added if services were offered to agencies which would assist them in meeting their primary missions in more effective ways.
Incentive A in the Task 5 report is based on the assumption that agencies would still be submitting publications for printing, and that GPO would then be making decisions about formats for the FDLP. This provides the opportunity for electronic conversion of publications, which could serve information needs of FDLP users and of the agencies themselves. But it may not provide a major incentive for agencies to move away from print altogether, unless the services offered by GPO can provide efficiencies beyond what the agencies can do in other ways.
Incentive B applies to information products which agencies do provide electronically and maintain themselves, and suggests ways to assure that the information is included in the FDLP. It addresses to some extent the need for a FDLP even when information is available somewhere on the Web. FDLP partners assist users in identifying appropriate and authoritative information, and provide sites for access by users who have no direct Web connections. With these benefits, and the added proposal that the GPO would assist in transferring electronic information as required by NARA, it could provide incentives for agency cooperation with the FDLP, for information already in electronic format. It is less clear whether these advantages would be enough to act as incentives to migrate additional information from print.
The issues identified in the Task 5 report are important ones which deserve additional prominence as the study progresses. The need for standardization, at least for a consensus on a group of acceptable formats and software, is expressed by both agencies and users. This process should provide a mechanism to move toward acceptance of standards, not to be imposed by GPO but to be agreed upon by all branches of government. GPO's service could be to evaluate alternatives and assist with implementation.
Task 5 concludes that even in the electronic environment there is a need for a central focus for coordinating public access to government information. If the government is to carry out its commitment to public access to its information, a central coordinating authority will be the most cost-effective way to assure that.
The problem with trying to use the FDLP as an incentive to move agencies to electronic publishing is that agencies have many responsibilities and many pressures on limited budgets. The provision of information to libraries is not the primary mission for most agencies, and their incentives to adopt electronic publishing must come from a broader vision of what will serve agency missions and also prove cost-effective. If elements of the FDLP and services offered by the GPO can be proven to assist agencies in these ways, participation in the program can provide a viable incentive for migration to electronic information dissemination.
TASK 6: Evaluation of current laws governing the FDLP
and recommendation of any legislative changes necessary for a
successful transition to a more electronic program.
ABSTRACT: The draft changes to Chapter 19, Title 44, aim to facilitate the transition to a more electronic program. Chapter 19 should be amended to recognize the electronic focus of the program and to ensure that the growing array of electronic products and services published by all three branches of government are included in the program. The entire life-cycle of information-- from its creation to its permanent access and preservation--and agency compliance are additional issues that need to be taken into consideration as legislative changes are considered by Congress.
Section 1. Scope of Information in the FDLP
The draft language definitions of "Government information," "Government information product," and "Government electronic information service" (1a) indicate that information produced in a variety of electronic formats, including both tangible products and online services, are as much within the scope of the program as materials produced in print formats. The Task 6 draft also suggests language that would bring into the program materials that have in the past been excluded. These include cooperative publications that must be sold by agencies in order to be self-sustaining (1b); fee-based electronic services (1c); and products not produced or procured by GPO (1e).
In the current budget environment, there is concern that agencies may impose copyright- like restrictions on government information products, both in print and electronic formats. Congress needs to address this issue as it conflicts with Principle 5, "Government Information Created or Compiled by Government Employees or at Government Expense Should Remain in the Public Domain." A stated goal of the GPO study was to find ways of using technology to improve and enhance the public's access to information. To be successful, the FDLP is dependent on Congress to provide sufficient funding, either directly to agencies or through the Superintendent of Documents, to make these materials available to the public at no cost.
As the number of agency electronic information products grow, the role of the GPO in providing users with bibliographic and long-term access becomes even more critical. A mechanism whereby the Superintendent of Documents is able to access electronic source data files from agencies is vital to ensuring that such data becomes a part of the program, is easily identifiable to the public, and is available for the long-term.
Section 2. Permanent Public Access to Government Information.
The proposed programmatic changes shift responsibility for permanent public access from participating depository libraries to the government. In view of the fact that agencies are today developing web sites with neither standards nor requirements for long-term access, a significant loss of valuable information is already occurring. The proliferation of agency web sites will exacerbate this loss unless legislative changes clearly define roles and responsibilities of all participants. Agencies should comply not only with making information available to the public, for example through an agency web site, but also with assuring that the files are transferred for permanent access to either the GPO or another archival facility. Legislative changes should consider the entire life-cycle of electronic information.
The draft language suggests that coordination by the Superintendent of Documents may accomplish the goal of permanent public access. Other than proposing use of GPO's electronic storage facility, however, the draft language lacks specifics as to which entities are to be ultimately responsible for permanent public access. More precise language would be useful. In addition, sufficient incentives, including funding, are necessary to entice program libraries to participate in a distributed system for permanent long-term access.
Finally, more precise recommendations are needed to address the preservation of data, migration of formats as necessary, distributed storage of data and equipment, and long-term public access concerns. Until these issues are addressed and resolved, any transition to an electronic depository program is incomplete and will result in a significant loss of access to government information by the public.
Section 3. Requirements for Depository Libraries.
Depository libraries in the past have fulfilled the requirement for providing public access and service with outstanding commitment. The transition to a predominately electronic program, however, imposes new and significant responsibilities and costs. It is questionable that the premise that each depository library, even small selectives, would be able to provide public access and service to all materials to which the locator service links. Assuredly, a program library must meet and probably exceed the proposed minimum technical guidelines in order to provide adequate public access. However, a program library should have the flexibility to provide expertise and service depending on their own user community needs and collection strengths. The draft language suggested to expand 44 U.S.C. 1909 is vague and not sufficiently specific to provide guidance for designation of program libraries.
Section 4. Notification.
It is important that the draft language notification requires that an agency inform the Superintendent of Documents when an information product or service is initiated, substantially modified, or terminated. This provision parallels the notification requirement of the Paperwork Reduction Act of 1995 and is necessary in order for GPO to provide bibliographic access and to coordinate permanent access to agency electronic information services. The notification requirement will enable GPO to provide full and timely bibliographic access to these products and services so that the public can derive the maximum benefits from the value of the information.
Section 5. Compliance Issues.
In order to meet the stated principles and goals of enhancing the public's access to information through the use of electronic products and services, legislative language is needed to ensure agency compliance in all three branches of government. Agencies must have adequate and positive incentives for participation in the program but there must also be penalties for non-compliance.
Section 6. Cataloging and Locator Services.
GPO's coordinating role of providing users with a catalog of Government information products and services, and with the locator service should continue. The success of these endeavors is directly related to whether or not agencies comply with the notification requirement. The public must be assured that the GPO cataloging and locator services are comprehensive and timely since these services will be a primary point of access to all electronic government information.
Section 7. Redescribing the Program to Reflect a Changing Environment.
The library community has long recommended that the FDLP program be renamed to become more meaningful to the general public. The suggested new language, "The Federal Information Dissemination and Access Program," was in fact introduced during the Chicago Conference on government information and more recently supported by the library associations in the Enhanced Library Access and Dissemination of Federal Government Information: A Framework for Future Discussion.
TASK 7: Survey Federal Agencies to Identify CD-ROM Titles
Not Currently Included in the Federal Depository Library Program.
ABSTRACT: Task Group 7 surveyed government agencies regarding their inclusion of CD-ROM products into the FDLP. Possible solutions to the problem of agencies' bypassing the FDLP with important CD-ROM titles are: improved communication with agencies; more precise language in Title 44 to recognize that electronic information falls within the scope of the program; and better cooperation between the agencies and the FDLP to ensure that software licenses are negotiated for FDLP libraries. It is very alarming to learn from the survey that over half of agency CD-ROM titles fall outside of the FDLP.
Task 7 addresses the need for empirical data regarding agency participation in the FDLP. It surveyed federal agencies to determine reasons for not including CD-ROM titles in the program. Responses to the survey indicate that the three most important reasons for non-participation in the FDLP were agencies' lack of understanding of the requirements of Title 44 as they apply to CD-ROMS; restrictions imposed on software licenses negotiated by agencies for their CD-ROM products; and lack of communication between GPO and the agencies concerning inclusion of their products in the program.
Unfortunately, none of the agencies who responded to the survey gave any specific reasons for participating or not participating in the FDLP. The survey concluded that 55.6% of agency CD-ROM titles were identified by agencies as not included in the program. This means that almost half of the CD-ROM titles are not readily available to the public at no fee at their depository library. Responses to the survey were also incomplete, making it difficult to make predictive and prescriptive statements based solely on this data. Because of this situation, the Task Group also used data obtained from ACSIS and compared it to the survey results to see if GPO has distributed any titles which agencies indicated were not included in the program.
Given the responses to the survey, better communication with the agencies regarding their responsibilities for making their CD-ROM products available to the FDLP is of paramount importance. Although the study recognizes that the language in Title 44 includes CD-ROM products, the definitions in sections 1901 and 1902 should be strengthened in order that agencies share this recognition. Software licensing is another area which should be addressed by both the agencies and by GPO. As Task Group 7 points out in its report, "GPO can (and has) contracted for software licenses for sales and depository copies when agency licenses do not cover GPO dissemination."
Fostering better communication between GPO and the agencies hinges on several assumptions, including the acceptance and recognition of the need for a central coordinating authority such as the FDLP to ensure dissemination of federal information products and services to the public through libraries. Furthermore, legislative changes to Title 44 would better enable agencies to include their CD-ROM products in the FDLP. Whereas the numerical data gained from the survey is instructive, even more interesting is the casual attitude taken by the respondents, both in some agencies' failure to respond to the survey and in the inaccuracy of some of the data provided. As the Task Group concludes, "a program of improved communication or outreach to agencies may be necessary to ameliorate this situation." As with other aspects of the study, implementing this conclusion is predicated on the assumption that adequate funding is provided to the program.
The issues raised by this task group become even more important as individuals and organizations are increasingly turning to CD-ROMs as a permanent solution to the problem of access to government information after its usefulness in the online environment or on the web has decreased.
TASK 8A: Evaluate the costs and benefits involved in converting
Congressional bills and resolutions to electronic formats for
distribution through the Federal Depository Library Program.
ABSTRACT: Alternative B eliminates microfiche distribution of Congressional bills and resolutions in favor of a monthly cumulative CD-ROM containing the PDF files. The option of selecting these important materials on CD-ROM would allow the public to access them in a cost-effective and user-friendly manner. The final annual cumulative version would provide libraries with assured access to older materials that might be withdrawn from the GPO server. Depository libraries would also have timely access to these important materials in PDF files through GPO ACCESS. It is important that Congressional bills and resolutions be accessible through mirror sites in order to provide the depository library community with a sense of security that online access to recent Congressional bills and resolutions would be available at all times.
The distribution of Congressional bills through the Federal Depository Library Program (FDLP) began with paper distribution. At the beginning of the 97th Congress in 1981, the distribution format for Congressional bills changed from paper to microfiche. Although there were concerns expressed about the suitability of microfiche for this important category of depository library materials, the switch to microfiche distribution enabled many libraries to more easily maintain collections of Congressional bills. A paper finding aid, arranged by category and then by bill number, provided a finding tool for locating the text of Congressional bills within the microfiche collection.
Today, the availability of online services and CD-ROM technology provides the opportunity to explore other avenues for dissemination of Congressional materials, including Congressional bills. These options have been explored in the report on Task 8A which had as its mission to, "Evaluate the costs and benefits involved in converting Congressional bills and resolutions to electronic formats for distribution through the Federal Depository Library Program."
The task force report states that Congressional bills on microfiche are selected by 859 depository libraries at a cost of approximately $94,940.00 per Congressional session. 544 depository libraries select the electronic version of bills available through GPO Access. Although 544 libraries officially select Congressional bills in electronic format, it is safe to assume that some depository libraries are making use of Gateway Libraries, or directly accessing the GPO World Wide Web site. It should be noted that it is currently possible to select both microfiche and electronic Congressional bills.
Alternative A: Eliminate all microfiche distribution to depository libraries and make Congressional bills and resolutions available online through the WAIS server. The PDF files for the bills could also be mounted for FTP download.
Providing online access to Congressional bills would enable those libraries that are technologically capable to benefit from access to current Congressional bills. Many depository libraries are fully equipped to access the GPO WAIS server via the World Wide Web and/or telnet; however, a large percentage of depository libraries are not technically capable of doing so. All depository libraries should be able to access this important source of public information. While the microfiche may be difficult to read and is not arranged strictly in numerical order, it is useable and patrons can access the materials they may need. While technologically-capable libraries may provide electronic access to current Congressional bills through GPO Access, how difficult will it be for a library to provide access to the older materials that will need to be withdrawn from the server because of space considerations? Will this interface be transparent for the user?
As noted in the disadvantages to this alternative, the Task Force stated that, "If depository access to historical files is to be ensured, a less costly and longer term distribution method will be needed to supplement online access to the bills. This may mean production of a CD-ROM or mounting of the PDF and ASCII files for FTP downloading after a predetermined period of time." Providing access to Congressional bills solely in an online environment will negatively affect the ability of many depository users to access both the current files of Congressional bills as well as retrospective files that may be housed at separate locations.
Alternative B: Eliminate microfiche distribution of the Congressional bills and resolutions in favor of a monthly cumulative CD-ROM containing the PDF files. Depository libraries would still be able to access the online service.
This alternative provides an economic and user-friendly approach to distributing Congressional bills and resolutions. It is estimated in the draft report that GPO would save approximately $34,032.00 if this approach to dissemination of Congressional bills were adopted. While saving costs, this approach would also provide a useful product with the ability to search and download the text of Congressional bills. At this time, libraries need to use other finding aids, often commercially produced, to determine the location of bills they need.
The 1995 Biennial Survey indicates that 83.1% of all depository libraries have CD-ROM capability. This percentage makes it reasonable to expect that depository libraries would select a CD-ROM product if it were available. A monthly cumulative CD-ROM will eliminate the filing and storage problems associated with the microfiche bills. In addition, if kept on a regular schedule, a monthly CD-ROM product would be more up-to-date than the current microfiche distribution, which has been subject to contractor delays. As Internet technology becomes more stable, and as depository libraries meet minimum technology requirements for participation in the FDLP, it may become unnecessary to produce a monthly update and an annual CD-ROM may suffice.
TASK 8B: Evaluate the costs and benefits involved in converting Congressional Documents and Reports to electronic format for distribution through the Federal Depository Library Program, even though currently a substantial amount of the source data is not available to GPO in machine readable form. (Attachment D-8)
ABSTRACT: The Congressional documents and reports have provided a significant, ongoing, historical record of the work of Congress. Both the bound paper version and the individual slip versions of this material has been distributed through the Federal Depository Library Program (FDLP). A combination of Alternatives A and C would provide comprehensive access to this valuable information. Depository libraries would have timely access to most of the recent documents and reports through GPO ACCESS; older materials would be accessible either through CD-ROM or the bound paper Serial Set. For the electronic product to be useful though, Congress needs to determine what it considers the authoritative version to be.
Task 8B is to evaluate the costs and benefits associated in converting Congressional Documents and Reports to electronic format for distribution through the Federal Depository Library Program. This effort must be examined in conjunction with the production of the U.S. Congressional Serial Set. It is important to keep in mind that these are two distinct series. The individual slip documents and reports are produced first; the bound Serial Set volumes are produced much later.
The U.S. Congressional Serial Set comprises a significant portion of the historical record of the work of Congress. The legal basis for compilation, binding, numbering and distribution of the paper bound Serial Set is contained in 44 USC sections 701, 719 and 738. The Serial Set currently includes Senate and House documents, congressional committee reports, presidential and other executive publications, treaty materials, and selected reports of nongovernmental organizations.
At present, every depository library is eligible to receive both the slip publications and the bound Serial Set in either paper and/or microfiche format. For the 101st Congress, 1st session, the cost to GPO for producing and distributing the Serial Set was $1,567,000. This figure covers 463 libraries receiving the Serial Set in paper and 755 libraries receiving microfiche.
The conversion of documents and reports to electronic format is problematic at present. While a high percentage of the reports are available in machine readable format, only 20% of the documents are received from Congress in this format. In order to be put online, GPO has to scan the materials to convert to a machine readable form. Unfortunately, this does not always work resulting in a non-searchable image file only. In order for this process to be effective, GPO will need to receive all reports and documents in machine readable format at the start. In addition, some documents are too graphic-intensive to ever be converted to electronic format.
With this in mind, the Working Group has proposed three dissemination alternatives in Task 8B. All three alternatives continue the production of a bound paper Serial Set, although alternatives B and C only allow regional depository libraries to receive copies. The Serial Set is a very important compilation and a key historical record to providing an ongoing collection of the publications of the U.S. Congress.
Alternative A: This option provides regional depository libraries with the bound Serial Set and the slip Documents and Reports through online access as well as in a CD-ROM version. Selective depositories could choose online access to the slips in lieu of either paper or microfiche. Selective depositories would also be able to select either the bound Serial Set or the Documents and Reports CD-ROM.
Alternative B: Alternative B provides the bound paper Serial Set only to regional depository libraries. Selective depository libraries would have the ability to select the Documents and Reports CD-ROM, which would be issued quarterly, cumulating for the session. All libraries would have the option of accessing the reports and documents online from GPO Access.
Alternative C: This option would supplement Alternative B by providing the option of distributing paper copies to depository libraries of any Documents and Reports too graphically intensive to practically convert to electronic format.
The value of this collection of Congressional materials is considerable. The Working Group may wish to consider a combination of A and C to provide optimum public access. All depository libraries that perceive a need for the paper bound Serial Set should be able to continue to receive it. The individual slip documents and reports, except for those too graphic-intensive, would be available online until the quarterly Documents and Reports CD-ROM is distributed to all libraries.
Whichever alternative is chosen to provide the slip documents and reports through the FDLP, there is one issue that Congress still needs to address--what is considered the authoritative version of the reports and documents? Will an online version be considered the authoritative version? Will the CD-ROM version? In conjunction with this issue is the need to guarantee the authenticity of the electronic version.
TASK 8C: Determine the costs and the impact on public access to the Department of Energy (DOE) technical reports through the FDLP as the Office of Scientific and Technical Information (OSTI) moves forward with its efforts to convert these reports from microfiche to electronic format. (Attachment D-9)
ABSTRACT: The Department of Energy's Office of Scientific and Technical Information (DOE/OSTI) is switching from microfiche production to a wholly electronic method of dissemination. The production of a fiche format is expected to end after FY 1996. At that time DOE/OSTI will be entirely electronic. The Department of Energy is committed to providing access to these valuable materials through theFDLP. Alternatives A and B propose making DOE/OSTI reports available through their Web site which would provide very timely access although because of the large number of image files, downloading would be very slow. Alternative C proposes access through CD-ROMs which would not be as timely and would require comprehensive cumulative keyword indexing. Depository libraries, particularly regionals, should have access to both formats with reliance on the Internet for the most recent reports, and on the CD-ROM for older materials.
DOE/OSTI materials have caused many libraries, especially Regionals, space problems due to the large number of microfiche sent each year. During FY 1995 17,117 unique reports were shipped out to those FDLPs that selected them. In discussions over the past few years depository libraries have tried to find ways to ease the burden of storing all of these fiche. Some suggestions have included having only a few libraries receive these materials and furnish copies to the rest of the system and another area that has been discussed, especially in Regional meetings, is to have fiche on demand, i.e., only provide fiche titles upon request from individual libraries. It was assumed that this would be less costly than providing large number of libraries with all of the titles. Having this material on demand electronically would solve all of the space problems and potentially make the reports more timely.
At the moment GPO and DOE/OSTI have a shared agreement that GPO pays only for the distribution costs for DOE reports. DOE pays for producing the fiche and for the depository copies. They also agree to fulfill missing publications claims and provide abstracts and indexing services for the reports (GPO does not catalog these publications or list them in the Monthly Catalog). The DOE/OSTI has been very cooperative in meeting depository library needs and has been a responsible agency in terms of participation in the program to provide DOE information to the widest number of users possible. The task force report states on page 2 of Attachment D-9 that DOE/OSTI is committed to providing access to DOE reports free of charge to depository libraries regardless of any policy decision they make concerning general public access. This is a most commendable public service position for the DOE to take and the Depository community appreciates their efforts on our behalf to ensure that we are included as a part of their information process.
This case study gives three dissemination alternatives. The first two, alternatives A and B, are virtually the same except for who pays for the costs. In these two scenarios DOE/OSTI allows depository access to the reports Web site. No fiche, paper copy, or CD-ROM would be available through the Federal Depository Library Program (FDLP). Cost savings would accrue to both agencies. Additional libraries would be able to serve the public with electronic access to this DOE Web site. The scenario further states that just-in-time access is provided instead of just-in-case access. In alternative A DOE/OSTI pays for the computer resources, user support, and depository library usage. In Alternative B incremental costs for FDLP usage would be paid for by GPO from their Salaries and Expenses appropriation. In both cases the study states that one disadvantage that users who access the Web site through a modem would have is trouble downloading because of the large size of the image files--a problem that Internet users would not have to the same degree although it too can be very slow. Also in both cases each agency might find increased costs due to unlimited usage.
The advantage to the FDLP is ready access to reports on a potentially more timely basis. Libraries would not have to provide long term storage for this material and the library would only obtain the titles that their patrons actually needed. The disadvantages would be the same as raised in other areas concerning on-line electronic material, i.e., increased costs to library for hardware, problems of downloading big files, abilities of library and patrons to use electronic information, and concerns over long term archiving and public access issues (which are not addressed in this case study). Also Internet access may require local software, i.e., Adobe Acrobat or something similar, to view documents and the depository libraries may also have to distribute copies of such software to users to take with them to read the material.
In Alternative C DOE/OSTI reports would be made available to the FDLP only on CD-ROMs and not on-line through the DOE Web site. These CDs would be packed with DOE reports in random order (DOE/OSTI estimates approximately 125 title per CD). GPO would premaster the CD-ROMs from DOE image files. A key benefit of this alterative is that depository libraries are better able to handle CD-ROMs than Internet sources (the 1995 Biennial Study shows 83% of FDLP have stand alone workstation with CD-ROM). Also CD-ROM access means that there is no reliance or strain on the DOE Web site (DOE experiences no additional loads on their computer resources) and extended access is provided all across the country at FDLPs. Downloading large image files would be easier on libraries using CD-ROMs than through a modem. The stated disadvantages are that CD-ROM access would not be timely, additional expenses would be incurred by GPO in creating and maintaining indexes to each CD, and those FDLPs that do not select the DOE CD would still have to rely on those that did. Also GPO would probably have to consider comprehensive cumulative keyword type indexing to compete with the quality of Internet access. Another consideration not addressed is that creation of DOE reports on CD-ROM would call for some software package to access and use the files on CD. Such a software system should be user- friendly and place no additional burdens on depository staff and hardware, nor impose any copyright-like restrictions. Also the library may have to provide copies of the software to their users in order to make viewing possible at home.
Another possible scenario not proposed in the draft report is that DOE/OSTI and GPO cooperate to extend access to depositories in both formats, especially to Regionals. This would give timely and current access to DOE reports through the Internet and would allow Regionals or some other selected group to select and house a less timely CD-ROM version for storage. This, of course, would be more expensive to the agency and/or GPO but would offer some choices to depositories, and make downloading of big files easier and faster. Perhaps some costs could be saved by offering Internet for current materials and CD-ROM access for older material.
Finally the case study leaves three issues not addressed. The first and foremost concern is the one that seems to bother depository librarians the most: that is there is no mechanism or policy to ensure extended, long term public access to a agency Web site or that the data will be maintained on any WWW site. If this problem were resolved and the FDLP was assured that this type of access would be guaranteed then the major arguments against Internet access could be laid to rest and libraries could get on with solving the hardware and access burdens that such electronic access causes them. Another concern is that Web sites are intended to serve the agency's major constituency, and providing public access through the FDLP places additional burdens on the agency's equipment, staff, and resources. If this burden is too great or has not been given a great deal of study by the agency, it could lead to a change of heart by the agency and result in restricted access or the imposition of user fees, etc. Last but not least, the study points out that agencies must understand that access through the FDLP means that their services should be designed for multiple simultaneous users from the same library without limitations such as single-user passwords.
TASK 8D: Identify issues that must be addressed when an agency no longer makes electronic information dissemination products and services available at its Web site, and the site contains information that needs to remain available to the public through the Federal Depository Library Program (FDLP) and/or transferred to the National Archives and Records Administration (NARA).
ABSTRACT: Task Group 8D recognizes that the Office of Technology Assessment (OTA) exemplifies the case study of an agency no longer maintaining its Web site (in this case due to the agency's demise); furthermore, in this particular situation, the Web site includes reports that have not been formally published. The task report affirms that agency Web sites, which may contain information not available in any other format, "...are in essence forms of publication and therefore may be Federal records as defined by 44 U.S.C. 3301."
The Task 8D report states that GPO is primarily interested in providing continued short-term access (5 years minimum) for much of the information on agency Web sites, while NARA focuses narrowly on that portion of the information which has historic value, with the goal of assuring preservation of that information. This is an oversimplification of the goals of the FDLP, since the Regional depository plan was developed primarily to guarantee permanent retention and access to the information distributed through the FDLP. Since the Task Force Report emphasizes shifting the responsibility for permanently maintaining and providing access to government information from depository libraries to the federal government, there is concern that: 1) some federal government information may "fall through the cracks" and eventually disappear, perhaps, for example, because it has not been saved in the GPO electronic storage facility, it has been lost in the transfer of data from one site to another, or it does not meet NARA's criteria for historic value; and 2) that there will continue to be adequate bibliographic control of this information for both retrieval and inventory purposes. Task 8D repeatedly addresses these important issues.
The report suggests two dissemination alternatives for GPO regarding OTA electronic files, with the understanding that: 1) OTA has already made arrangements to mount information from OTA Online on GPO's Web site; and 2) OTA also has a contract to scan all the texts of their reports dating from 1972 and convert to Acrobat PDF format; these files will be packaged along with much of the information available via OTA Online and some additional historical material on a set of five discs. Alternative B, which would have the OTA CD-ROM set distributed to depository libraries, and after a predetermined period of time, OTA information would be removed from the GPO Web site, is more cost-effective and has fewer disadvantages/problems than Alternative A, which has GPO maintaining the OTA information on its Web site as well as distributing the CD-ROM collection upon completion, with no plan for permanent retention of the OTA files. The Task Group may wish to consider a third alternative which effectively combines Alternatives A and B, but has GPO transferring the OTA files to NARA for permanent retention, after the CD-ROM set of OTA reports has been completed and distributed to depository libraries. This would eliminate the problem of NARA not accepting the CD-ROM set because it uses the PDF software-dependent format, and also would allow NARA to accession only those files which were unique or of historic value, knowing that a complete set of files was available through the FDLP.
Regarding the appraisal alternatives, Alternative A, which would have NARA accession the records of the persons/committees responsible for maintaining agency Web sites, with the idea that these records would reflect the content and structure of the site, is less satisfactory than the other alternatives offered. The 8D report admits that "This option...ignores the possibility that in the future, the information posted on the Web site might not appear in any other format...[so] it is necessary not only to appraise the records of those maintaining the files, but the files on the Web site itself." This is a real situation; the FDLP already has begun to distribute federal information solely in an online format.
While Alternative B, which has NARA accessioning all files within a Web site, is more comprehensive than Alternative C, in which NARA would accession selected files, there are potential problems involved in documenting the huge amount of files and links within some agency Web sites. However, there are also problems with Alternative C in which NARA would determine which files may not exist in any other format as well as which files have historic value, in order to decide which files to eventually accession.
One of the major issues identified in the 8D report is permanent FDLP access to electronic information dissemination products and services. The report asks "If information already has been distributed in paper, microfiche or CD-ROM does it make sense to provide continued online access to the information?" Yet in Alternative B in the OTA scenario, where the CD-ROM set of OTA reports would be distributed to depositories and the OTA information would be removed from the GPO Web site, it is considered a disadvantage for public access to the reports to be available only at or through depository libraries. GPO and NARA should work closely together to determine the best method of ensuring permanent FDLP access to government information. The concept of transferring responsibility for permanent retention/access from depository libraries to federal agencies may need to be revisited with the intent to consider compromises that fall between the two extremes. For example, one possibility might be for the FDLP to establish Regional electronic depositories which would be responsible for storing and providing access to information contained on federal agency Web sites; NARA would be able to select only those files considered to be unique or of historic value for retention in the National Archives.
The Task Group has brought an important issue to the forefront, especially since the report also states "If an agency decides to discontinue access to information through their Web site, does GPO have a responsibility to obtain the information and provide funds and resources for its continued access through the FDLP?" Since NARA is not mentioned in the discussion of this issue, there certainly is an implication that either depository libraries and their patrons (the public) should not necessarily expect to obtain access to this information through the National Archives, or, depositories and their patrons will not necessarily find the information as easily located and retrieved from NARA as it is through the FDLP. GPO and NARA should consider all of the federal information needs of the American public in order to determine the best arrangement the two agencies can work out between themselves and among all federal agencies to ensure permanent public access to electronic federal government information.
TASK 9: Evaluation of issues surrounding inclusion in electronic formats of materials not traditionally included in the FDLP in either paper or microfiche. Examples include: Federal district and circuit court opinions (Task 9B), SEC filings (Task 9A), patents, military specifications and a variety of other scientific and technical information (primarily contractor reports).
ABSTRACT: The Working Group is to commended for evaluating alternatives for improving access to these valuable materials through the Federal Depository Library Program. The materials considered in the Task 9 report have generally not been distributed through the program and yet the information clearly meets requirements for depository distribution. Cost considerations and other factors have restricted its dissemination through the FDLP although other similar material is distributed. It would enhance public access and be extremely useful to make STI (scientific and technical) data available electronically through the program. However, the imposition of copyright-like restrictions on the electronic dissemination of this data is very problematic.
The types of information considered in Task 9--patents, military specifications and standards, Congressional Research Service Studies, and scientific and technical information such as EPA technical reports and guidelines, DOD technical reports and NTIS reports--include resources of enormous importance to scholarly and industrial research and development. It is very helpful that the Working Group evaluated several alternatives for improving access to these materials through the FDLP. The materials considered in the Task 9 report have generally not been distributed through the program. Many are similar in nature to report literature, such as Department of Energy and NASA reports, which have been part of the FDLP. Patent literature has been available through a separate and more limited patent library depository program. It would be highly desirable to improve access to patents, specifications and standards, CRS Studies, EPA and DOD technical report literature through the FDLP. The information available clearly meets requirements for depository distribution; cost considerations and other factors have restricted its dissemination through the FDLP although other similar material is distributed.
Voluminous materials such as specifications and standards, patents, and STI (scientific and technical information) seem ideally suited to on-demand electronic delivery because of the costs and space required to disseminate, house and maintain either a paper or a microfiche collection. Any given report, specification or patent may be used infrequently, although the cumulative use of the collection may be high.
Not addressed in the Task 9 report is the issue of bibliographic access to these voluminous collections of STI materials. Increasingly, print indexes are being discontinued and are not necessarily being replaced by improved electronic versions. In the case of NTIS, its primary catalog and index is now privately produced and is not available in an electronic version at no cost. By contrast, the Patent and Trademark Office is greatly improving access to its materials through online electronic indexing and abstracting. In order to avoid losing our national research heritage, the cumulative results of millions of dollars of investment of public and private funds, maintaining both bibliographic access and access to the print or electronic versions of the documentation itself is important.
A major obstacle to FDLP dissemination of these valuable resources is the cost-recovery basis under which some agencies operate. Ideally, agencies should be funded to a level to permit no-fee distribution, at least to depository libraries, and to make charges to others based on the incremental cost of dissemination. In an electronic environment, such considerations have led agencies such as NTIS to propose the imposition of copyright-like restrictions on electronic dissemination of data. Relatively few NTIS publications are popular enough to sell enough copies to turn a profit and it would be possible for competitors to skim off and sell their own copies of popular titles. NTIS and other agencies are also concerned that if a depository library made an electronic publication freely available, the agency's own market would be negatively affected. Similar fears of negatively impacting the market for print or microform materials have not materialized. The proposal outlined by NTIS would impose copyright-like restrictions on the use and manipulation of government information.
Dissemination alternatives: In evaluating alternatives for dissemination, it should be assumed that no one alternative is appropriate for all the types of information discussed under Task 9. Also, it is critical that long-term access to and preservation of printed and electronic information be ensured. Alternatives C and D, which involve the Government Printing Office in the distribution process, would provide long-term access. Similar guarantees should be assured for any alternative selected.
Alternative A and B: Alternative A provides that agencies would make their own information available for dissemination through the Internet, at no cost to the user. The GPO Locator would direct users, including depository library users, to the agency site. Alternative B is similar, except that agencies would charge a fee for their information and GPO would negotiate an agreement to pay the costs of online access for depository libraries. The agreement could include limitations on number of users or on remote access via library networks, but would not include copyright-like restrictions on use or re-use of information.
Alternative A and B may be appropriate for voluminous data such as patents and information under the custody of NTIS or DTIC. Both alternatives would greatly improve access to materials which have never been available through the FDLP, and in both instances, the FDLP would provide assistance to users in locating and using the data. It is also true that displaying and printing extensive documents with tables and graphics will not be easy, and both libraries and end-users will need to acquire appropriate equipment, software, AND experience in making this information accessible. Even when information is disseminated at no fee, the costs to users will be significant.
Among the disadvantages of both alternatives would be that public access will put additional loads on agency computing and telecommunications resources as well as on support services. Nearly 1400 libraries could be potential users and would need access training and support.
In the current budgetary environment, it is unrealistic to expect that Congress will elect to completely subsidize the Internet dissemination of patents and STI. Thus no-fee access through the FDLP would be a substantial improvement in public access. Other data, such as specifications, are currently available at no cost and should continue to be, since electronic distribution may be a more cost-effective alternative for the agency.
Alternative C: This option provides that GPO would establish a database of information from agency sites which is tailored to the FDLP. This alternative would relieve agencies of concerns about unauthorized access to other information in its files, as well as the user load on its systems. It would also provide a desirable redundancy of access, maintaining availability of data in the case of damage at another site. Because of the voluminous nature of some of this information, it may not be economically feasible for GPO to create and maintain a separate database. However, for less extensive materials from agencies with security concerns, this alternative could be ideal. Under Alternative C, the standard interfaces GPO could offer, and the additional bibliographic access it might provide, would be important contributions to effective use of the information.
Alternatives D and E: In Alternatives D and E, GPO would distribute information downloaded from online sources to the FDLP in CD-ROM format, either produced by agencies (D) or GPO (E). CD-ROM distribution is the least desirable alternative, for a variety of reasons: the time delay in distributing the CD-ROMS; the sheer number of CD-ROMS that would need to be distributed; the difficulty in locating the required data on the CD; and the inability to update material distributed in CD-ROM format. Long-term access to these materials must be ensured. At present, CD-ROMS may offer an edge in terms of long-term access, but they do not provide the kind of on-demand access that may be more appropriate for large collections of data in which any given title receives little use.
Alternative F: This option was proposed by the National Technical Information Service after the completion of the Task 9 report. It is a variation on Alternative B, in which the information is available from an agency site, for a fee, but without the involvement of the Government Printing Office. It is a unique model in that valuable materials would be made available to the public for the first time through depository libraries, and yet the materials would not be an official part of the FDLP. The NTIS proposal requires an agreement from participating libraries not to release the electronic file outside the library or use it for commercial purposes. Such a restriction is necessary, according to NTIS, to assure that depository access and use do not infringe on the agency's own market. At the same time, this in effect amounts to a copyright-like restriction on the downstream use of these materials and would put librarians in the position of having to limit or even police the use of these materials.
On one hand, this overture from NTIS should be viewed as an opportunity to make important STI materials more readily available to the public through depository libraries. On the other hand, the proposal places restrictions on the use of government information that are expressly prohibited in Principle 5 of the draft report and indeed in the Paperwork Reduction Act. Of concern with the NTIS proposal is that it might become an accepted model for other electronic government information services. Therein lies a grave danger to the public's no-fee access through the FDLP. It is a serious issue which requires Congressional study and review.
Regarding the NTIS proposal, it would be useful for the pilot project to be carefully developed with input from the depository library community and the NTIS Advisory group. This is a very important undertaking that will add valuable materials to the program. Libraries will have a great deal of work to do doing the pilot project to establish mechanisms for printing documents. The pilot project should be useful for testing mechanisms of delivering material electronically to individual users that would not damage NTIS's market.
TASK 9A: Evaluate issues surrounding inclusion of the Securities Exchange Commission (SEC) EDGAR System in the Federal Depository Library Program when the information is not already included in paper or microfiche format. (Attachment D-12)
ABSTRACT: It is commendable that the SEC has taken full advantage of WWW technologies to provide no-fee access to the EDGAR database, a valuable public resource to company records. Task 9A proposes two alternatives for public access to EDGAR through the FDLP: the first suggests using the GPO Locator service to enhance the public's ability to access EDGAR through the Internet but does not address the need for multiple mirror sites nor the long term need for ready access to historical EDGAR information; the second, the distribution of CD-ROMs, may resolve the multiple site access and long term storage issues but would add expense and rely on a technology that may soon become outdated. Both alternatives have merit but a combination of both may be most desirable. A third alternative could be considered, not to replace the others, in which libraries, community civic networks, library consortia, and other not-for-profit organizations form partnerships with federal government information producing agencies. These partnerships will assure ready and timely access to EDGAR resources through redundancy of access to the information, as well as long term preservation of this important information.
The 9A Task Group has selected the SEC EDGAR System as an model of using the Internet to increase public access to electronic information. The commitment of SEC Chairman Arthur Levitt to resist pressure to privatize the EDGAR System and post it directly to the WWW serves as a model for other federal information providers. Appropriately, with the advent of Internet access, Chairman Levitt has concluded that the SEC has the responsibility to make these materials equally available to the public--individual users, libraries, and the private sector.
In the draft study, Task 9A describes two alternatives for providing access to SEC EDGAR information. In the first, access to the EDGAR system would be strictly online; GPO and depository libraries would incur little expense. It is assumed that GPO would add value through sophisticated indexing in its Locator service which would be used by the public, libraries, and private sector information businesses alike. The FDLP ensures that the knowledge and skills of government information specialists are available in all Congressional districts to assist and train members of the public unfamiliar with accessing federal information. In this alternative the SEC and the public derive significant value from GPO indexing and depository library assistance at very little expense.
A key concern with Alternative 1 is long term access to EDGAR records. The FDLP has traditionally guaranteed long term access to federal publications through regional depository libraries. Alternative 1 suggests no mechanism that will assure the ready availability of government publications that have been provided through regionals. Though the SEC is engaged in negotiations with NARA to schedule retention of EDGAR materials, we are concerned that access to archived federal information is less immediate through NARA than it is through regionals.
A second key concern raised in Alternative 1 is redundancy of access--that is, the availability of access through more than one source in the event that the primary channel (in this case the SEC) is interrupted. Given current Internet capacity and technology, disruptions of service are not uncommon. The stability of individual systems is also at best uncertain, as typified by system crashes and power failures. This option provides no alternative for accessing EDGAR data in the event that the SEC data platform is incapacitated or regions of the country are unable to connect via the Internet to SEC databases.
Alternative 2 proposes the tangible distribution of SEC data to depository libraries on CD-ROMs and provides a possible solution to both problems of long term access and lack of redundancy. By depositing EDGAR data on CD-ROMs in regional depositories--or some other sub-set of depository libraries--complete sets of EDGAR information will be available at no-fee from multiple sites. These libraries would accept their traditional responsibilities for maintaining the information and providing it to the public either directly or through other depository libraries. As major players in the increasingly electronic information universe, they would bear the responsibility for migrating the data to new media as information storage technologies evolve. In this way, multiple sites would provide long term access to EDGAR information resources. We recognize that this alternative incurs potentially significant expenses. However, the value added by these costs in terms of the free flow of federal information to the public warrants the investment.
A possible third alternative would be the establishment of partnerships between the SEC and individual libraries, library consortia, library associations, community networks, or other not-for-profit organizations. In such partnerships the partner libraries would operate under agreements with the SEC to serve as no-fee mirror sites for the EDGAR database; provisions for long term access would be included. The federal agencies responsible for guaranteeing public access to federal information, such as the GPO, NARA, and OMB, would provide guidance and coordination in drawing up such partnerships. In this alternative the value of EDGAR is still guaranteed to the public but at little expense to the federal government. Partner libraries would accept this responsibility as a part of their mission and service to their constituencies, and with the understanding that many other libraries are embarking on similar arrangements to provide no-fee access to other federal, state, and local government information resources.
Overall, Alternative 2 provides needed dependability and resolves the important questions of long-term and redundant access associated with the strictly online scenario proposed in Alternative 1. However, the increased expenses associated with Alternative 2 may suggest that additional models, such as that of partnerships with no-fee mirror sites, be explored.
TASK 9B: Evaluate how United States Court of Appeals' published slip opinions might be included in the Federal Depository Library Program (FDLP) electronically, although they have not been a part of the FDLP in either paper or microfiche format.
ABSTRACT: United States Courts of Appeals' slip opinions have not previously been included in the Federal Depository Library Program (FDLP). We believe that incorporating the electronic version of these slip opinions into the FDLP is consistent with the view of the Senate, expressed in Senate Report 104-114, that "advances in technology provide new opportunities for enhancing and improving public access" to Government information. The development of depository access should be based on new and emerging Internet technologies, and not on the outdated bulletin board systems which are rapidly becoming obsolete. In order to provide an electronic product that would be useful to the public, any option selected must be able to guarantee the authenticity of the opinions and ensure the provision of long term access to this essential public information.
In a letter dated February 16, 1996, the American Association of Law Libraries (AALL) provided comments on the Task 9B report which investigated the possibility of including U.S. Courts of Appeals' slip opinions electronically in the Federal Depository Library Program (FDLP). We appreciate the fact that you took these comments under consideration and were very pleased to see that many of them were incorporated into the latest draft Task 9B report.
The Courts of Appeals' slip opinions have not, to this date, been included in the FDLP. Incorporating electronic slip opinions into the FDLP is a perfect example of the use of advances in technology to "provide new opportunities for enhancing and improving public access" to Government information. (S. Rep. No. 114, 104th Cong., 1st Sess. 48 (1995)). Our associations endorse the inclusion of the slip opinions in the FDLP as a very positive step towards realizing the Senate*s goal of improved public access. Although Task 9B is limited to U.S. Courts of Appeals" slip opinions, we believe that it should serve as a model to provide the public with electronic no-fee access to the opinions of the Federal District Courts as well.
Whichever alternative is ultimately selected to provide electronic slip opinions through the FDLP, there are two important issues that need to be addressed. The first is authenticity. A means of guaranteeing the authenticity of the electronic version is essential. Law is a discipline which relies on precedent. Legal researchers, including legislators, attorneys, law students and faculty, and the general public, should all be assured that the information is both reliable and the most current authoritative version. The second issue is preservation and long term access. In Section 4, the draft report raises certain questions that need to be addressed yet it fails to suggest any answers. We believe that, as technology advances, the public has the right to a seamless transition from the slip opinion to the final authoritative electronic version. In addition, the government has the responsibility to ensure the permanent availability of the final authoritative version, at no cost to the public, and in a format that will be usable with future technologies, as current software and hardware become obsolete.
We affirm the position expressed in the February 16 letter that options B and C are not viable. Both of these options rely on bulletin board systems (BBS), a model that has several disadvantages. First, BBSs use a technology that is rapidly becoming obsolete. In contrast, the Internet alternatives offer the advantages of speed of transmission and full text searching. Second, the BBS model is decentralized and lacks a single standard setting authority. With no central authority, the slip opinions are likely to suffer from a lack of standardization as it applies to file formats as well as search and retrieval software. In addition, this lack of standardization inhibits verification of authenticity and complicates preservation efforts.
The following comments on Alternatives A, D, and E are in addition to those expressed in the letter of February 16, 1996.
Alternative A: GPO ACCESS
The success of this option, to provide slip opinions through GPO ACCESS, is dependent upon changes to Title 44 which would require the courts to supply GPO with the electronic slip opinions. Although the Courts of Appeals have historically been granted a waiver from the requirement to use the printing services of the Government Printing Office (GPO), such a waiver is not necessarily appropriate in an electronic environment, and would inhibit any efforts to provide comprehensive access to all of the slip opinions through the FDLP. In addition, to be effective, any such change to Title 44 must include adequate enforcement provisions. The use of GPO ACCESS would meet the Congressional goal of improving and enhancing public access to government information as long as GPO ACCESS remains available free of charge to the public. In addition, the GPO ACCESS option would provide one centralized standard setting authority in GPO. Preservation and long term access will however, depend on continued long term funding of the GPO ACCESS system by the Congress.
Alternative D: Judiciary Web Site
This option, to provide slip opinions on the Judiciary web site, would be an improvement over the current bulletin board systems since one central standard setting authority, presumably the Administrative Office of the U.S. Courts (AO), would be established. While this option would certainly be a technical improvement over the current decentralized system of BBSs, which we consider to be obsolete, no-fee public access must be ensured. Again, preservation and long term access will depend on funding and a commitment on the part of the AO to guarantee maintenance and archiving of the opinions.
Alternative E: Consortium of Law Schools
The efforts of the law schools which provide Internet access to the slip opinions are notable because the consortium is committed to making them available to the public free of charge. Although this model is decentralized, there is evidence of law school cooperation (e.g., in the development of keyword searching across sites) that could be expanded to include standards for authenticity, preservation and long term access. Ultimately however, preservation and long term access will depend on the continued efforts of each individual law school.
TASK 10A: Review the effects of offering free public access to STAT-USA information products and services through the Federal Depository Library Program (FDLP). (Attachment D-14)
ABSTRACT: STAT-USA, a cost-recovery service within the U.S. Department of Commerce, produces business and economic information products, including the Economic Bulletin Board (EBB), the National Trade Data Bank (NTDB) on CD-ROM, and STAT-USA/Internet. These products are available through the Federal Depository Library Program (FDLP), and are among the most heavily used electronic government information sources at depository libraries.
This task report articulates the dilemma, from an agency's perspective, of trying to balance the competing mandates of cost-recovery and wide public dissemination. STAT-USA is the product of an agency that recognizes the value of including its information in the FDLP. But while the agency has cooperated with GPO to provide its products to the public through depository libraries, public access is restricted by current practices and pending changes to pricing and access policies.
The federal government should adequately fund public access to government information resources produced for public use, and must address the role of cost-recovery programs in the menu of public information dissemination services. Competing laws mandate, in some cases, that agencies both provide no-fee access to the public through depository libraries and at the same time recover costs for those same services. In addressing this apparent dilemma for self-funded agencies, Congress should, at minimum, reaffirm the public's right to no-fee access to government information through the FDLP. A broader public deliberation of the issues of "fee vs. no-fee" must take place in order to clarify these difficult policy and technical issues.
The list of tasks for the GPO study task groups included the key issue of fee-based services in the tenth task: "A review of Federal programs permitting or requiring the sale of information to recover costs, and the effects on efforts to assure free public access through the FDLP." This task addresses a central policy question, where emerging technologies are providing both opportunities for broader access and problems in identifying and recovering costs. While there are many examples of programs which fall into this category, the study includes only two case studies, STAT-USA (Task 10A) and MEDLINE (Task 10B). Since this is such an important and complex issue, it is one which requires additional data and consideration by Congress to reconcile conflicting policies and assure appropriate support for programs which carry out the government's information principles.
As a case study, the STAT-USA program is an excellent illustration of the problems faced by an agency which operates in a fee-based environment and yet wishes to provide a level of access to its materials through the FDLP. In trying to adapt the FDLP model for tangible products to the electronic environment, STAT-USA is facing the difficulties inherent in controlling the use of electronic information, which is easily networked and shared and hard to contain. Because the products from STAT-USA are enormously useful to FDLP users, the libraries want to provide the broadest possible access.
Carrying the traditional FDLP model into the electronic age is more complicated than it may seem at first. For example, the NTDB CD-ROM includes on it about 250,000 publications. Many of these represent materials which were formerly in the FDLP in paper. For each of these publications, a depository library received one copy without charge; if it wanted more copies, it could purchase them. At any one time, multiple users might be reading many of these multiple publications in a depository library. If the users wished to have their own copies of materials, they could buy them from government sources or pay for photocopies in the library. As printing ceases and publications are bundled onto the NTDB, the depository library still receives one copy without charge, but now it has thousands of publications on one CD. Unless the CD is placed on a network, the number of possible simultaneous users of these many publications is cut down to one. From the library point of view, networking of the NTDB provides a level of access similar to that provided in the paper environment, but from the agency point of view it could cut into the sales which are necessary to sustain the product.
The development of STAT-USA/Internet introduces additional issues. The Internet product is not identical to the CD-ROM. There are many time series and matrix tables on the CD which are not online, and these are of major interest to the research community, and thus should remain in the FDLP. But the Internet STAT-USA provides timely access and consistent searching, significant advantages for many FDLP users. The establishment of the Internet version has presented the agency with the challenge of registering users and controlling their use of the information they receive. The administrative problem of registering depository libraries was solved by the cooperation of GPO's Library Programs Service, which took over that responsibility, and that cooperative model deserves replication for other agency Internet services which might be added to the program.
Controlling the use of information is more problematic, and introduces the issue of asking libraries to enforce copyright-like restrictions on the use of government information which go beyond any controls libraries needed to impose on the use of tangible formats. Users have always been free to photocopy paper and fiche publications, and use the copies without restriction. Electronic dissemination provides the opportunity for much easier and broader redissemination, and this could undermine the relationship between publishing agencies and the FDLP. With paper and microfiche formats, no-fee use in depository libraries was not a serious threat to the sale of materials for individuals, organizations and businesses which wished to have the convenience of their own copies. STAT-USA is trying to replicate that model with the provision of one free password for use in each depository library, but the libraries are anxious to provide access to more than one user at a time through networking. Only one person is some congressional districts with only one depository library would be able to access this materials at any one time under this proposal. Since the Internet offers the opportunity to provide public access to government information when and where it is needed, the government needs to come to grips with the issue of support for that broad and beneficial access.
This same issue was faced by the GPO itself, which like STAT-USA had statutory language which permitted charging reasonable fees (for users other than depository libraries) for its online GPO Access system. After more than a year of experience with maintaining complex registration procedures and charging non-depository users for access, the GPO decided to make the entire system free to all users. The resulting changes in use and in costs and revenues for the GPO would provide useful additional data and should be incorporated into this study.
The two alternatives presented in the Task 10A report seem to imply that the NTDB CD-ROM would remain in the depository program since the contents are not all covered in the Internet version and the CD-ROM provides long-term access for information not included in the Internet version. The only real difference between the two alternatives is where the funding for the costs of Internet access would come from. Alternative A would fund depository access from other STAT-USA fees, since the agency no longer has sufficient appropriated funds to support FDLP participation. This might seem similar to universal service in the telecommunications field, where all users pay to support basic service for those who would not otherwise have it. It would succeed only if libraries could limit redissemination so that the income which supported the program was not destroyed, a delicate balance indeed.
Alternative B acknowledges that there is a cost to providing FDLP access to STAT-USA, and proposes that GPO would pay for depository access through its appropriated funds. This option includes some cost figures which may have been superseded by more recently-released fee schedules from STAT-USA, which propose higher fees for networking both the CDs and STAT-USA/Internet. The proposal to have GPO pay for FDLP access to fee-based government information services appears in several of the task reports and also in the GPO's Strategic Plan, but there seems to be little data on the actual costs which this might incur. From the user's point of view, the essential issue again is that the government should fund adequate public access to the information resources for which the American public has already invested.
The "Issues to be Addressed" section of this task group articulates the difficult problems of funding public access and the "fee vs. no-fee" controversy. Depository librarians see the great variety of uses made of data provided through STAT-USA, and are convinced that the public benefits from the broadest possible transfer of economic information, to new and established businesses as well as to students and researchers. To make such information totally fee-based would be contrary to the principles set forth in the GPO study.
The challenge faced by agencies, the Congress, and depository libraries is to develop a new model for access to electronic government information, which will continue to provide the public with access to government information which is mandated in Title 44 and reinforced by many other statutes and directives. The tension and even conflict between statutes which require access and those which require cost recovery is exacerbated by new technologies, even as those technologies provide opportunities for more efficiency and better access. More deliberation of these vital public policy issues is necessary.
TASK 10B: Evaluate alternative for including the National Library of Medicine (NLM) MEDLINE data, available as an electronic fee-based service, in the FDLP. (Attachment D-15)
ABSTRACT This Task Group has bought together GPO and NLM for a serious discussion of the issue of providing depository libraries with access to MEDLINE. Further discussions should take place regarding NLM's proposal for a pilot project with a limited number of depository libraries. Since Grateful Med is now available through the Internet, that option should be explored further. Costs of providing this access can be more accurately assessed after a pilot activity.
There currently exist many access points for health sciences librarians, health professionals, health sciences students, and historians to use the library's resources. All hospitals and medical schools offer access to MEDLINE and other database resources and Grateful Med is designed specifically for the end-user searcher. In addition, many public libraries offer CD-ROM or other access to these files. The transition to an electronic environment in this case might well involve an examination of existing offerings of this information and may well present depository libraries and the FDLP the opportunity to explore cooperative arrangements with NLM for services and training.