vol21.2 58 IASSIST Quarterly The Electronic Freedom of Information Amendments of 1996 [1] In September, 1996, the U.S. House of Representatives and the U.S. Senate passed the “Electronic Freedom of Information Act Amendments of 1996” on a voice vote, with no debate, and with support from both Republicans and Democrats. President Clinton signed the E-FOIA bill into law in early October. The E-FOIA amendments revise the text of the Freedom of Information Act (5 U.S.C., sec. 552) or, as it is commonly known, “the FOIA,” by addressing the subject of electronic records for the first time. Many of the amendments took effect after a 180-day period, on March 31, 1997. Others do not take effect until November 1, 1997, and still others at a later date. While the FOIA and its E-FOIA amendments pertain solely to U.S. federal government records, the influence within the U.S. of this law is such that the ramifications of these new amendments can be expected to be closely watched among electronic records creators, providers, and users in both the public and private sectors of U.S. society, and potentially elsewhere. The purpose of the Freedom of Information Act, as enacted in 1966 and amended subsequently in 1974 and 1986, is to “require agencies of the [U.S.] Federal Government to make certain agency information available for public inspection and copying and to establish and enable enforcement of the right of any person to obtain access to the records of such agencies, subject to statutory exemptions, for any public or private purpose.” When he signed the E-FOIA amendments into law, President Clinton noted the important role FOIA had played in the previous 30 years in establishing an effective legal right of access to government information. He underscored the crucial need in a democracy for open access to government information by citizens. He offered his hope that as the Government uses electronic technology to disseminate more information, there will be less need for citizens to use FOIA to obtain government information. The legislative history prepared by the [U.S.] House Committee on Government Reform and Oversight as background for the E-FOIA amendments identifies the purpose of the amendments as providing for “public access to information in an electronic format, and for other purposes...” Their history highlights several key aspects of the E- FOIA amendments, from the legislators’ perspective: -- electronic records: the amendments make explicit that electronic records are subject to the FOIA. Furthermore they acknowledge the increase in the Government’s use of computers and encourage federal agencies to use computer technology to enhance public access to Government information. — format request: with implementation of the E-FOIA amendments, requestors may request records in any form or format in which an agency maintains the records. Also, agencies must make a “reasonable effort” to comply with requests to furnish records in [any] formats specified by the requestor, “if the record is readily reproducible by the agency in that form or format.” This change reverses a legal opinion that dates from 1984 and which formed the basis for federal agency practices since that time. This change will be discussed further, below. -- redaction: agencies redacting electronic records (deleting part or parts of an electronic record [or an electronic records file] to prevent disclosure of material that is exempted from release), shall indicate the amount of information deleted on the released portion of the record, unless doing this would harm an interest protected by the exemption. Further, “if technically feasible, the amount of the information deleted shall be indicated at the place in the record where such deletion is made.” — expedited processing: the amendments establish that certain categories of requestors would receive priority treatment of their requests if failure to obtain information in a timely manner would pose a significant harm. The first such category are those who might reasonably expect that delay in obtaining the information could pose an imminent threat to life or physical safety of an individual. The second category includes requests made by a person(s) primarily engaged in the dissemination of information to the public, e.g., the media, and involving a compelling urgency to inform the public. -- multitrack processing: the writers of the amendments An Early Perspective on the “Electronic Freedom of Information Act Amendments of 1996” by Margaret O. Adams* Summer 1997 59 created an incentive for requestors to submit narrowly specified information requests to federal agencies by allowing agencies to establish procedures to process FOIA requests of various sizes on different tracks rather than on a first-received, first-responded-to order. The assumption here is that requests for specific or smaller amounts of information can be completed quickly, so responses to such requests no longer need to be in a queue with more general or larger-volume requests. -- agency backlogs: in an effort to ameliorate the phenomenon of significant backlogs in responding to FOIA requests in many federal agencies, the amendments stipulate that agencies can no longer delay responding to FOIA requests because of “exceptional circumstances” if such circumstances simply result from a predictable agency request workload. — deadlines: the amendments extend the deadline for agencies to respond with an initial determination to a FOIA request to 20 working days, from the previous deadline of 10 working days. From the perspective of the executive branch of the government, the part of the federal government that has to implement the E-FOIA amendments, the effects of this bill are highlighted somewhat differently. According to the Justice Department’s newsletter, FOIA Update, a major change of the amendments concerns the maintenance of electronic access in agency reading rooms. Prior to the E- FOIA amendments, agencies were required to make three categories of records routinely available for public inspection and copying: final opinions rendered in the adjudication of administrative cases, specific agency policy statements, and administrative staff manuals that affect the public. The amendments add to the categories of reading room records and also establish a requirement for electronic availability of reading room records. The new category of records that agencies have to make available in their reading rooms as of March 31, 1997, includes any records processed and disclosed in response to a FOIA request that “the agency determines have become or are likely to become the subject of subsequent requests for substantially the same records.” By the eve of the new century, December 31, 1999, agencies are to have a general index to FOIA-released records and are to make this index available by computer telecommunications. Theoretically the idea is that making records in greatest demand accessible in an agency reading room should satisfy most future demand for those records. But, even for federal agencies that maintain public reading rooms in their regional offices throughout the country, this expectation may not be met. It suggests that the bill drafters assume that most of the public’s demand for records under FOIA can be satisfied by having an agency reading room where researchers can come to “read” records. Yet the amendments stipulate that anytime an agency receives a FOIA request for records, the agency must treat the request in the standard FOIA fashion, regardless of whether it also makes these records available in its reading room or online. In other words, even though it already makes such records available in an agency reading room or online, it must respond formally to the requestor within a 20-day time period, and provide the records under whatever guidelines and fees it has established for processing requests under FOIA. Note here: a FOIA request is any request for records that invokes or mentions the FOIA, or Freedom of Information Act. The E-FOIA amendments, as suggested above, also expand the concept of an agency reading room to what some are referring to as “electronic” or virtual reading rooms. The amendments require that agencies use electronic information technology to enhance the availability of their reading room records. And, for any “newly created reading room records [i.e., records created on or after November 1, 1996 that are in the category of “reading room records”], agencies must, by November 1, 1997, make these records available to the public by electronic means. Preferably this new electronic availability should be via computer telecommunications, i.e., in the form of on-line access, such as from a World Wide Web site(s) established to serve as “electronic reading room(s).” While the amendments do not explicitly state that agencies are to continue to maintain their conventional reading rooms, the advice in the Justice Department’s FOIA Update, is that they are to do this. In other words, the three categories of administrative and policy records traditionally maintained by agencies in their public reading rooms, plus any records released under FOIA that the agency determines are likely to become the subject of subsequent requests, must be made available for public inspection and copying in the agency public reading room. Further, any of these reading room materials created after November 1, 1996 must also be made available to the public by computer telecommunications by November 1, 1997. Finally, there are two additional new requirements that may have significant implications for agencies seeking to comply with both the spirit and the letter of the E-FOIA amendments. The first has already been mentioned earlier: agencies are to make records available in any form or format requested by the person if the record(s) is(are) readily reproducible by the agency in that form or format. The second new requirement was somehow not highlighted in the legislative history. Yet compliance with it may require substantial reorientation in the way federal agencies treat FOIA requests for information in federal records, when those records are in an electronic format. This is the requirement that states: “an agency shall make reasonable efforts to search for the records [responsive to a request] 60 IASSIST Quarterly except when such efforts would significantly interfere with the operation of the agency’s automated information system.” “Search” is defined as “to review, manually or by automated means, agency records for the purpose of locating those records which are responsive to a request.” Clearly, the impetus for the E-FOIA amendments just described comes from evolution in the use of electronic computer technology by both government agencies and the federal records-seeking public. It also reflects growing expectations for public access to more and more government information that has accompanied the proliferation of computer technology, especially personal computers. Some aspects of the amendments, however, reflect more than natural evolutionary change. The new right for requestors to choose the format in which they expect to receive federal records reverses long-standing legal opinion. The requirement that agencies use computer technology to search for records in electronic form reverses widespread federal practice rooted in a series of court rulings. To understand how or why these changes came to be law, it may be helpful to consider the historical context from which they emerged. “The Freedom of Information Act in the Information Age...” Our colleague Tom Brown recently published an article that examines the historical context of the FOIA. He discusses the case law on how FOIA related to computerized records prior to enactment of the E-FOIA amendments.[2] He notes that FOIA guarantees any person the right to gain access to records unless the records contain information on matters specifically excluded under one or more of the nine exemptions identified in the FOIA. Further if a portion of a record is exempt from disclosure then a reasonably segregable portion of the record is to be provided after deletion of the portions which are exempt. Brown notes that in interpreting the FOIA statute, the courts have consistently ruled that agencies are not required “to create records in order to respond positively to a FOIA request,” i.e., to provide records in response to a FOIA or to segregate exempted portions of records in order to release them in response to a request. Even prior to the E-FOIA amendments, the courts seemed to have resolved the question of whether electronic materials were subject to the FOIA. In a 1982 case cited by Brown, a federal appeals court ruled that “[C]omputer- stored records, whether...in the central processing unit, on magnetic tape or in some other form, are still records for the purposes of the FOIA.” Further, a 1989 Department of Justice survey of federal agency practices found that government-wide practice also affirmed that electronic records may be records under the FOIA. The question of whether the FOIA required federal agencies to provide requestors with records in the formats of the requestors choosing had generally been decided in ways that allowed the federal agencies to make that choice. For example, Brown discusses a 1984 case, Dismukes v Department of the Interior. The case centered on a FOIA request to Interior’s Bureau of Land Management (BLM) for a list of participants in oil and gas leasing lotteries in California. Dismukes, on behalf of the National Wildlife Federation — a private organization, had filed a FOIA for these records and requested that they be provided on 9- track, 1600 bpi magnetic tape, in an IBM-compatible format, and with file dumps and file layouts. BLM’s Office of Surface Mining provided computer printouts of the requested records. The National Wildlife Federation appealed this response, arguing that “it is impossible to work with such volumes of data without having it in computer form.” The Court of Appeals ruled that “the computer printout was fully responsive to the...request.” As a result of this ruling, the Department of Justice advised federal agencies that they, not requestors, had the right “to choose the format of disclosure, so long as the agency chooses reasonably.” This same principle was upheld a few years later when a requestor appealed the response of the Central Intelligence Agency (CIA) to a FOIA request for an index of documents that the CIA had previously released. In response, the CIA provided 5000 pages of printouts, arranged by date of release of the item in the index. In its appeal, the requestor asked that this information be made available on tape or disk. The court ruled however, that the “information [the CIA had provided] was in a “reasonably accessible form” and furthermore, that the agency was not obligated to provide in electronic format, records it had already provided in paper copy.” These rulings, Brown suggests, provide evidence that the courts were in denial of the computer age. While the courts may have been in denial, or in ignorance of the computer age, Brown also makes clear that their rulings were predicated on the basic premise that agencies are not “expected to be private research firms, ...subject to every beck and call of a requester.” To bolster this assessment, he cites a number of rulings. One, by a federal district court in Pennsylvania in 1988, clarified that FOIA does not require agencies to write new computer programs to search for data not already compiled for agency purposes. Another, a federal appeals court, determined that “the FOIA dictum to release reasonably segregable portions [of records] does not require creation of a ...summary file because it is not functionally analogous to manual searches” for records that contain information responsive to a request. Another concluded that the FOIA “in no way contemplates that agencies, in providing information to the public, should invest in the most sophisticated and expensive form of technology.” On the issue of format, Brown also shows that as early as 1988, the Administrative Conference of the United States (until October 1995, the federal government agency Summer 1997 61 responsible for studying and recommending improvements in administrative procedures to executive branch agencies) had recommended that in responding to FOIA requests, “agencies should provide electronic information in the form in which it is maintained or, if so requested, in such other form as can be generated directly and with reasonable effort.” Taken together, these rulings of the courts and the practices of some agencies responding to FOIA requests for information in electronic formats point to the rationale for the E-FOIA amendments. So, now, what can be their anticipated impact? Application of the E-FOIA Amendments In general, it is far too early to know how federal agencies will adapt their practices to be in compliance with both the spirit and the letter of the E-FOIA amendments. Similarly, it is much too soon to have any idea whether the cumulative effect of compliance with the E-FOIA amendments will result, as President Clinton said he hoped, in less need by citizens to use FOIA to obtain government information. There are a few things that we can suggest at this time however. J. Timothy Sprehe, writing in the Federal Computer Week (January 6, 1997) suggests that the principal benefits of the law “lie in the fact that EFOIA overturns two bad court decisions.” One of these was a ruling that the National Library of Medicine’s on-line information systems were not agency records for the purposes of FOIA. The other was the Dismukes case discussed above, where the court had ruled that an agency had the prerogative to decide the format in which it fulfills a FOIA request. But beyond this, Sprehe does not consider the E-FOIA to be “a great leap forward,” except in the sense that the “fact that it was passed at all may be an important reminder that the public has rights of access to electronic as well as paper-based information resources.” Another article, this one in a trade newspaper, Washington Technology (April 24, 1997), quotes an analyst for the Federation of American Scientists as saying “agencies must undergo a cultural transformation to accommodate [to] the requirements of [E-FOIA]...the law does not change reality...but it provides an incentive to modernize.” What he is referring to was further enunciated in this same article by a Washington lawyer who raises the question of whether federal agencies have the hardware, software, and personnel that will enable them to be in compliance with the E-FOIA amendments. As he states, the amendments “raise the issue of the availability of suitable software and the equipment to handle it, such as a client-server system with sufficient storage capacity and database software with advanced features. It also raises the issue of being able to hire and find personnel who are sufficiently trained to use what is sophisticated search and retrieval software to comply with requests.” The principles that underlie the FOIA and now the E-FOIA amendments are firmly based on the principles of the U.S. Constitution and its Bill of Rights. Yet, the IASSIST community, sophisticated and knowledgeable as it is in matters regarding maintenance and access to electronic records and information, might well ponder the implications for federal government agencies seeking to be in compliance with the FOIA as amended. Keep in mind that in general, the data community that IASSIST represents, offers or supports services for researchers knowledgeable in the structure and use of electronic data. Social science researchers and those who offer support services to social scientists, such as data archivists and librarians, are among those most likely to expect increasing expansion in access to electronic government information. Yet the need by social scientists, generally, for access to administrative and programmatic databases to use as sources for rigorous research and analysis purposes, are quite different than the needs reflected in a significant portion of the requests that, for example, the Center for Electronic Records at the U.S. National Archives and Records Administration, receives. Since our holdings reflect the records of the entire federal government, we can assume that at least in some senses, the requests we receive mirror those received by all federal agencies. In our case, for the first six months of the current fiscal year, approximately one-third of the inquiries (over 1100 — and they generated over 1800 separate responses) requested specific items of information from records in our holdings. Very few of these invoked the FOIA, an indication perhaps, of the well-known practice of the [U.S.] National Archives and Records Administration (NARA) to treat all requests as if the requestor had invoked the FOIA, thus negating the need for requestors to use it in order to receive the records or the information in records that they seek. That is, it is the policy of NARA to respond to all inquiries in a timely manner, and as responsively as possible. The extent to which it can successfully do this is in large measure a reflection of how informed the request is — how specifically it identifies the information sought, and whether the manner in which the specificity of the request reflects the description that NARA has for the relevant records. In our particular case, most requests for specific records or for information from specific records, pertained to the casualty records from the Korean and Vietnam conflicts, two of our best-known electronic records files. We have long experience in responding to such requests, and the E- FOIA amendments will have virtually no impact on the manner in which we handle responses to these inquiries — from printouts of the full files, or by using some pc-based versions of the files, with retrieval software, that a small business vendor has created and provided to us in beta-test 62 IASSIST Quarterly format. Two individuals have developed web sites with these records where anyone can access them. Interestingly, such widespread availability of these records seems to have had no impact on the continuing demand that we receive for specific information from them. In microcosm at least, this experience suggests that the ready availability of records does not stem the direct demand for them. So, our real challenge in complying with the E-FOIA requirement to search electronic records for information responsive to a request will not be in regard to the casualty records. Rather it will be in responding to requests for information that may reside in any of the other 30,000 (and growing) electronic records files in our holdings. And this is but a reflection of the challenges facing federal agencies as a whole. While each agency only receives requests for information from its records, whereas NARA receives requests for information from archival records of the entire federal government, the records in agencies are more current and thus they are in more urgent demand than archival records usually are. There is no arguing amongst us that technological innovation makes access to public information more efficient and varied in ways that none of us could have imagined even just a few years ago. But, we also know too that with each innovation, new complexities and possibilities have also presented us with the reality of a seemingly infinite rise in the level of expectations for what we can do or offer. Keeping pace with such expectations, or rather, determining the nature of the “reasonable” effort by which we measure the limits for responding to those expectations, is perhaps the greatest challenge for those seeking to provide service to citizens that is responsive both to the spirit and the law embodied in the FOIA, as amended. The experiences of the data community can assist and influence these determinations. Outreach by the data community to the larger universe can perhaps also help to inform their expectations. NOTES 1. The first portion of this paper is based primarily upon the Committee on Government Reform and Oversight, U.S. House of Representatives Report 104-795, 104th Congress, 2d Session, Electronic Freedom of Information Amendments of 1996: Report [to accompany H.R. 3802], September 17, 1996. This paper also draws upon U.S. Department of Justice, Office of Information and Privacy, FOIA Update, Vol. XVII, No. 4, Fall 1996. 2. Thomas Elton Brown, “The Freedom of Information Act in the Information Age: The Electronic Challenge to the People’s Right to Know,” American Archivist, Vol. 58, Spring 1995, pp. 202-211. Much of the analysis in this section of this paper draws upon Brown’s article. * Presented at IASSIST-97 (May 9, 1997), Odense Denmark Name: Job Title: Organization: Address: City: State/Province: Postal Code: Country: Phone: FAX: E-mail: URL: IASSIST INTERNATIONAL ASSOCIATION FOR SOCIAL SCIENCE INFORMATION SERVICE AND TECHNOLOGY ASSOCIATION INTERNATIONALE POUR LES SERVICES ET TECHNIQUES D'INFORMATION EN SCIENCES SOCIALES I would like to become a member of IASSIST. Please see my choice below: Options for payment in Canadian Dollars and by Major Credit Card are available. See the following web site for details: http://datalib.library.ualberta.ca/iassist/ mbrship2.html $40 (US) Regular Member $20 Student Member $70 Subscription (payment must be made in US$) List me in the membership directory Add me to the IASSIST listserv Membership form The International Association for Social Science Information Services and Technology (IASSIST) is an international association of individuals who are engaged in the acquistion, processing, maintenance, and distribu- tion of machine readable text and/or numeric social science data. The membership includes information system specialists, data base librarians or administrators, archivists, research- ers, programmers, and managers. Their range of interests encompases hard copy as well as machine readable data Paid-up members enjoy voting rights and receive the IASSIST QUAR- TERLY . They also benefit from reduced fees for attendance at regional and international conferences sponsored by IASSIST. Membership fees are: Regular Membership. $40.00 per calendar year. Student Membership: $20.00 per calendar year. Institutional subcriptions to the quarterly are available, but do not confer voting rights or other member- ship benefits. Institutional Subcription: $70.00 per calendar year (includes one volume of the Quarterly) ○ ○ ○ ○ Please make checks payable, in US funds , to IASSIST and Mail to: IASSIST, Assistant Treasurer JoAnn Dionne 50360 Warren Road Canton, MI 48187 USA http://datalib.library.ualberta.ca/iassist/mbrship2.html http://datalib.library.ualberta.ca/iassist/mbrship2.html R eturn U ndelivered M ail T o: IA S S IS T Q U A R T E R LY c/o W endy T readw ell 1758 P ascal S t. N orth F alcon H eights, M N 55113 U S A