Author Archives: dcornwall

We have had a chance to review the new Congressional Research Service (CRS) Report Federal Depository Library Program: Issues for Congress (Petersen) available at from the Federation of American Scientists, Project On Government Secrecy web site.

While we believe it serves as a useful overview of the Federal Depository Library Program (FDLP), the report has a few significant problems. Members of Congress should consider the following before using this report as a basis for modifying the FDLP:

Report appears to take Ithaka S+R report at face value

Pages 6-11 of the CRS report concern the findings of the Government Printing Office (GPO) commissioned Ithaka S+R FDLP Report (Housewright) and GPO's ultimate rejection of the report. We are concerned that CRS has taken Ithaka's conclusions at face value and have not considered the many criticisms of the Ithaka report. Some of these criticisms included:

• The report made broad statements about users without sufficient consultation with actual end users.
• The report focused on the value of the program to libraries and not to users.
• The report apparently ignored corrections from law librarians and others so that errors in draft documents carried over to final documents.
• The report excluded serious discussion of digital deposit and local digital collections of federal information.
• The report failed to account for risks of implementing its recommendations.

We wrote extensively during the Ithaka S+R report period. We were not alone. A complete set of comments that Ithaka S+R received on its project web site is available from GPO. Yet the CRS report authors do not appear to have considered the public comments that questioned a number of Ithaka S+R's findings.

Another curiosity is CRS's omission of GPO's reasoning for rejecting the Ithaka S+R report. The authors simply note that "GPO did not provide a detailed, publicly available explication of its decision." It seems to us that it would have been appropriate and useful for CRS to have contacted Superintendent of Documents (SuDoc) Mary Alice Baish and interviewed her about GPO's rejection of the Ithaka S+R report. In doing so, CRS could have expanded the existing public record with more details from GPO as to why the report was unacceptable and given the report an additional depth of understanding. Given the GPO rejection of the Ithaka S+R report and the amount of criticism of the report from the library community, CRS's reliance on the report results in a description of the FDLP that is both limited and slanted.

Threats to access to digital government information

Pages 13-14 of the CRS report address "Access to Digital Government Information." The section concludes with the following:

The use of the FDLP Electronic Collection may raise the following concerns in the context of digital information:

• Where do FDLP Electronic Collection data reside?
• Are current data management protocols sufficient to ensure no loss of data availability, and assured access?
• Are those protocols similar in GPO, other federal agencies, and non governmental partners that provide content?
• What backup, and information distribution and assurance policies, are in place?

Although these are legitimate questions, CRS left out bigger problems within which these questions are merely details of implementation. These bigger problems stem from the GPO-centric model of the FDLP in which GPO has usurped from libraries the roles of both preservation and access. By replacing libraries, GPO has endangered the long-term future of information preservation and free-public access to that information in many ways. Three of the most important of those are, uncurated access, the term we call "silent withdrawals," and the very real potential of inadequate funding of GPO along with the complementary danger of replacing of free access with fee-based access.

Uncurated access

Access to government information has been a key tenet of the FDLP for 200 years. CRS averred that fact when they stated, "emergence of digital delivery of government information outside the FDLP program may offer increased access to government information to those who might not be able to visit depository libraries." But the key point missed by CRS is the idea of uncurated access. By only discussing access, but not preservation, CRS ignores the processes carried out by depository institutions to *preserve* govt information. We have said many times on FGI that access today does not equal access in the long-term. Libraries have begun to put processes in place to assure long-term digital access (University of North Texas Digital Library, LOCKSS-USDOCS, Archive-it collections, End-of-term crawls etc). Librarians can and should continue their curatorial responsibilities in the digital realm. We can't expect GPO and other government agencies -- especially in this budget crisis climate -- to have the long-term vision necessary to assure long-term preservation. Curation and content control will be key issues going forward. These issues were merely glossed over by the report.

Silent Withdrawals

One of the many strengths of a distributed depository system is the way its very structure protects information from intentional or unintentional loss, censorship, or erasure. Without this protection, information can too easily be withdrawn "silently" -- that is, without public announcement or review. That FDLP works is evident when one compares information in the depository system to information not in the depository system. The number of documents that have been sent to depository libraries and later withdrawn is relatively small and the reasons for the recalls are usually not controversial.

• Federal publications recalled from libraries. (1981-2010). Stanford University Library.

Contrast this to information that has been withdrawn from the web, reclassified by agencies, and documents that have had open access restricted by agencies after their release:

• Less Access to Less Information By and About the U.S. Government (1981-1998).
• Chronology of Disappearing Government Information (1998-2002), Compiled by Barbara Miller for ALA/GODORT Education Committee With special assistance of Karrie Peterson.
• No URL Left Behind? Web Scrub Raises Concerns, by Michelle R. Davis, Education Week, 22:3 (9/18/2002), 1-2.
• Homefront Confidential. Reporters Committee for Freedom of the Press, Sixth edition (September 2005)
• The Age of Missing Information. By Steven Aftergood, Slate (March 17, 2005).
• Army Blocks Public Access to Digital Library, by Steven Aftergood, Secrecy News (February 13th, 2008).
• DOD withdraws embarrassing report, FreeGovInfo (2009-05-13).
• Secrecy Report 2011. OpenTheGovernment.org.

The reason for the success of the depository system is that it has checks and balances and procedures that must be followed when an agency wishes to withdraw a publication ("ID 72" GPO 2005). In the world of physical deposit of print documents, withdrawal of a previously deposited document requires the compliance of tens or even hundres of libraries that actually have physical possession and control of copies. While depository librarians have a legal obligation to comply with withdrawal and destroy orders, there have been cases where this step triggered complaints about unreasonable withdrawal requests. Such questioning has led agencies to withdraw requests that seemed based on embarrassment or paranoia rather than error or true security needs.

One noteworthy example of this comes from 2001 when the CIA put pressure on the Department of State to destroy already-printed volumes of the Foreign Relations of the United States, 1964-1968, V. 16, Cyprus, Greece, and Turkey. But those volumes were in the possession of GPO and slated for deposit with FDLP libraries (Aftergood). The volumes were not destroyed and were distributed (S 1.1:964-68/v.16).

Another example comes from 2004 when the Justice Department demanded that depositories destroy copies of five publications that dealt with, among other things, how citizens can retrieve items confiscated by the government. The American Library Association objected, the Justice department rescinded its order, and GPO allowed libraries to keep copies and also replaced copies already destroyed. (Lee)

GPO's policy does have good procedures to prevent "silent withdrawals" even of information that is not physically deposited with libraries. But when GPO does not deposit digital copies with libraries, depositories are cut out of the procedures and an important safeguard is missing. Withdrawal decisions and their execution stay wholly within the federal government -- making it easier for the government to remove items from public access. The "LOCKSS-USDocs" private LOCKSS network project is beginning to replace this safeguard, but more work is needed to ensure digital deposit with more libraries in order to guard against silent withdrawals.

Budget Problems

The current GPO-centric model of digital access described, and apparently unquestioned, by CRS has a single point of failure. If Congress decides it is no longer worthwhile to adequately fund information dissemination in general or GPO in particular, users and libraries will lose access to material unique to GPO's servers. Even the maintenance of so-called "persistent" URLs (PURLs) could be endangered by something as simple as inadequate funding.

Digital information requires long-term, consistent funding. Neither digital information preservation nor access can be accomplished passively: both require constant attention and renewal and resources. Even budget cutbacks can cause loss of information or loss of access to information. The single-point-of-failure GPO-centric model of preservation and access is a system in which even inadequate funding means loss of information.

Reduced funding can also lead to privatization of government information access. This can occur if the fee-based private-sector takes over the delivery of services that GPO drops because of inadequate funding. It can also occur if Congress mandates that GPO use a fee-for-service model. In both cases, free access will be lost and people and libraries may be unable to afford adequate access. (Jacobs)

An April 10, 2012 Federal Times demonstrates that GPO is already feeling a lot of pain:

At risk of needing a congressional bailout 18 months ago, the Government Printing Office slashed its workforce, cut employee benefits, rented out excess office space and took other steps to stabilize its finances.

To make ends meet, GPO is also focusing on money-making activities like making secure credentials for the FBI. At its heart, the FDLP is a cost center. It has no opportunity to make GPO profit. This is right and proper, but will continue to make the FDLP a tempting target in future budget reductions. (Jacobs)

Summary

Any discussion of disruptions in user access needs to acknowledge the above facts. As long as digital storage is centralized in GPO, free and permanent access is only a Congressional Act away from being disable or terminated. The report does ask a key question: what solutions might create a more robust FDLP that is better equipped to meet the demands of providing government information to American citizens." We at FGI and many allies in the FDLP community have been working on that question (see Letter to Deputy CTO Noveck: "Open Government Publications," Rethinking the Cloud, and Achieving a collaborative FDLP future to contextualize the issues involved).

The report written by Petersen, Manning and Bailey provides a useful historic overview of the FDLP. We feel that it somewhat mischaracterizes recent efforts at building consensus. Most seriously, the report leaves out major barriers to free, permanant public access to government information that MUST be addressed in any meaningful reform effort.

References:

• Aftergood, Steven. State Dept Mulls "Book Burning." Secrecy News (September 21, 2001)
• Housewright, Ross, and Roger C. Schonfeld. Modeling a Sustainable Future for the United States Federal Depository Library Program’s Network of Libraries in the 21st Century, Final Report of Ithaka S+R to the Government Printing Office, Ithaka S+R, May 16, 2011, Including "Statement from the U.S. Government Printing Office" by Mary Alice Baish (August 5, 2011)
• Jacobs, James A. Privatization of GPO, Defunding of FDsys, and the Future of the FDLP, FreeGovInfo (2011-08-11).
• Lee, Christopher. Justice Dept. Rescinds Order to Pull Publications. Washington Post (August 5, 2004) page A17.
• Lipowicz, Alice. Digital information may strain GPO and library system, CRS says, Federal Computer Week(Apr 11, 2012)
• Petersen, R. Eric, Jennifer E. Manning, and Christina M. Bailey. Federal Depository Library Program: Issues for Congress by Petersen, Congressional Research Service CRS Report R42457 (March 29, 2012)
• Reilly, Sean. Cost-cutting saves GPO from financial crisis, chief says. Federal Times, (4/10/2012).
• Replacement copies of five DOJ documents for depositories that had removed them. Judith C. Russell, Superintendent of Documents (August 5, 2004)
• U.S. Government Printing Office. Managing Director, Information Dissemination, Superintendent of Documents. Information Dissemination Policy Statement 72 (ID 72). "Withdrawal of Federal Information Products from GPO’s Information Dissemination (ID) Programs." Effective Date: June 21, 2005, Supersedes No.: SOD 72 Dated: 07/22/02.
• LOCKSS-USDOCS private LOCKSS network: The Digital Federal Depository Library Program.

Continue reading →

Continue Reading →