Analysis of the Regional Discard Policy: What you need to know about implementation

August 1, 2016 by
Filed under: Library, post 

Now that GPO is testing the implementation of implementing the Regional Discard Policy, it is time to finally get answers to the important questions that GPO has not addressed over the last two years.

Although we have asked many questions about this policy, our concerns boil down to one simple question: Will FDLP — and more importantly, the public! — lose information when Regionals discard their historic collections? To answer that question, we just need to know if GPO’s implementation of the Policy will ensure that the paper preservation copies of discarded volumes and their digital surrogates are complete and accurate.

We now know that the implementation will not prevent loss of information.


In October 2015, GPO announced that it would begin testing processes and procedures of the new Regional Discard Policy with six regional depositories in January of 2016. GPO provided little description of how the policy would be implemented beyond the policy itself, leaving many important questions unanswered.

GPO has now provided links to three draft implementation documents, a recording of an information webinar, and some other “informational resources.” This provides the FDLP community with the first indications of how GPO is interpreting its new Policy and how it intends to implement it.

There are still many unanswered questions, but we now have the first solid indications of how GPO intends to balance preservation of (and access to) the FDLP Historical Collections while discarding them.

We analyzed GPO’s earlier implementation statements in a previous post. We also asked specific questions about implementation in an open letter to GPO, and, in a separate post, we review the few answers to those questions that we can infer from this new batch of documents.

In this post we focus on the big picture and the most important things FDLP librarians need to understand about GPO’s implementation of the Policy.

GPO interprets "Access" to mean "Preservation"

When the Joint Committee on Printing (JCP) — the committee charged with oversight of GPO — approved GPO’s Discard Policy, it imposed one condition that GPO had not asked for: that FDLP retain at least four tangible copies, distributed geographically, for the purpose of meeting GPO’s needs but also for "providing the necessary access to the materials."

We now know how GPO is interpreting this. According to the Memorandum of Agreement (MOA), these copies will not be used for “providing necessary access” as JCP directed; they will instead be considered "Preservation Copies of Record." GPO defines these as a "version set aside to protect its informational content and intrinsic value from decay or destruction" stored in a "Controlled Access Environment … in which library users cannot directly browse and pull material from library shelves."

Paper preservation copies may be defective

The MOA defines the "Preservation copy of record (for tangible publications)" as being "The average used and worn publication that has all pages or leaves present." But GPO does not describe how (or who) will determine if all pages or leaves are present. The MOA also says that "any defects must be noted" — which presumably means that incomplete and damaged copies can be considered "preservation copies of record." Indeed, the procedures seem to anticipate that stewards will have damaged paper copies: A regional library that wishes to discard a title is instructed to send a copy of that title to the Preservation Steward if the discarding library’s copy is "in better condition than that of the Preservation Steward." But, again, the current implementation documents provide no procedures or instructions or standards that a library could use to compare copies or determine which copy is "better."

Impermanent Preservation of paper copies

We also know that those "Preservation copies of record" will be "preserved" by reliance on temporary contracts. GPO’s replacement of the legislative mandate for regional libraries to preserve the Historic Collections permanently is a “Memorandum of Agreement” that can be cancelled at any time. These MOAs between GPO and "Preservation Stewards" will specify that each steward will store one copy of a specific list of titles that are eligible for discard. The MOA "expresses" a "commitment" to permanent preservation, but does not obligate the stewards to a permanent (or even long-term) commitment. In fact, the agreement explicitly allows the steward library or GPO to terminate the agreement at any time by giving 120 days written notice. According to the FAQ, the agreements will even be reviewed every three years "for continuation." If a steward library ever wants to cancel the contract for preservation, it can do so without asking permission of anyone.

GPO makes preservation promises it cannot keep

The MOA says that, if a steward library cancels its agreement to store paper copies covered by the agreement, GPO "will find another depository library" to store those documents or will store those documents itself. In saying this, GPO contradicts itself and makes a promise that, by its own reasoning, it will likely not be able to keep. On the one hand, GPO has insisted that libraries need to discard paper copies and on the other hand it is promising that it will be able to find an unknown number of libraries (depending on how many libraries cancel their MOAs) to house an unknown number of documents — after those very documents have (presumably) been discarded. What makes GPO think this will be easy to do or even possible? GPO’s promise to store the documents itself contradicts another part of the MOA which says that GPO does not have a tangible collection and therefore it has to rely on preservation stewards for preservation storage.

Less Access: No Inter-Library Loan of paper copies

We have found no mention of Inter-Library Loan in the documents provided by GPO, but the MOA does limit use of the "Preservation Copies" to "in-building use only or for use in a special collections reading room." Presumably, this will prohibit those copies from being loaned through ILL. (The MOA also says that "The copy of record may be nondestructively digitized to create a digital surrogate to function as the use copy" but it is not clear why this is necessary since there is already supposed to be a digital copy in FDSys for that purpose.)

No guarantee that digital copies will be complete or accurate

GPO and others have argued repeatedly that the Discard Policy is reasonable because people will use the digital copies in FDsys. This argument becomes meaningless if the copies in FDsys are incomplete or inaccurate.

Unfortunately, although the Policy asserts that those FDsys copies will be complete, it provides no procedures for determining either completeness or accuracy — though we already know beyond a doubt that such digital surrogates contain errors and missing pages (Conway). The MOA has a single requirement for the "digital copy of record": it must be "produced to specifications that will allow the creation of a printed facsimile version." Separately, the "Instructions for Regional Depository Libraries to Request Approval to Discard" specify that the digital copy in FDsys must be "complete (no missing pages)," but, again, GPO provides no procedures to ensure such completeness. (The Instructions do specify that, for a title to be added to the list of titles eligible for discard, a library must compare the digital copy to the tangible copy to ensure they "match." But GPO’s instructions for this comparison only focus on determining if the two are “the same document,” not if the digital copy is complete.) Neither is there any requirement to determine if the digital copy is accurate. This is a significant omission because most of the volumes that are the targets of the Discard Policy ("more than one million titles" according to Vance-Cooks) will be digitized copies of print-originals. The best study we have so far (Conway) estimates that 35% of such titles are not accurate and complete enough to be considered “reliably intelligible surrogates.” (See also The Digital-Surrogate Seal of Approval: a Consumer-oriented Standard.)


The above are facts. If you read all the documents, you also get an impression of how GPO is implementing the Discard Policy so far.

One impression we get is that GPO is bending over backwards to make it easy for Regionals to discard documents and for libraries to sign on as Preservation Stewards. If the requirements for discarding and for providing stewardship included procedures that would ensure against loss of information content, it might be harder to discard and more difficult to attract “stewards.” This means that, although GPO is keen to express its goal of preservation and access, it is not translating that goal into actions that will ensure either.

Second, the risk of loss of information with these implementation procedures is high. Consider the fact that most of those one million titles that Regionals are eager to discard are born-analog print-originals. The only way those volumes will become eligible for discard is for someone to digitize them and for GPO to ingest them into FDsys. What standard has GPO set for such digitizations? The lowest possible standard: page-images. What procedures and standards is GPO requiring to ensure those digitizations are complete and accurate? None. As noted above, the most reliable study available tells us that more than one third of such digitizations are not complete and accurate. GPO’s announced procedures are a recipe for discarding complete and accurate publications and replacing them with inaccurate and incomplete digital copies. Libraries that moan that they do not understand why users do not value them should ask themselves how their users will react when they learn that we could have prevented this loss, but instead chose to create it.

Third, we are dismayed that the procedures are legally temporary. Nothing prevents libraries from cancelling their MOAs and GPO’s procedures for dealing with that are not only weak, they are demonstrably unrealistic.

Finally, we have to look to the future. The current GPO administration admirably declares that long-term preservation and access are essential parts of its mission. But, if even this GPO administration cannot or will not take the actions necessary to uphold those values, what will happen when a new administration with less admirable values takes over? When this administration can essentially ignore JCP instructions to provide paper copies for access and when it can use temporary contracts to provide permanent access, what will a less enlightened GPO do in the future? Given that GPO is bending over backwards to accomodate the demands for “flexibility” from a few libraries, and rushing to do so without adequate research, testing, or data, and without safeguards in place to prevent loss of information, what will keep a future GPO administration from abandoning any of these lip-service “commitments” by simply cancelling the MOAs?

Since the implementation is (presumably) still being tested and since the implementation documents provided are drafts, what we know so far is still preliminary. There is still time for the FDLP community to stand up and demand better. The next opportunity to do so is the fall 2016 Depository Library Council meeting and conference. FGI will be there. Will you?


Authors: James A. Jacobs and James R. Jacobs


4 Comments on Analysis of the Regional Discard Policy: What you need to know about implementation

  1. Hallie Pritchett on Mon, 1st Aug 2016 11:43 am
  2. Isn’t the entire FDLP “legally temporary”? Nothing prevents libraries – including regionals – from relinquishing their depository status entirely at any time ( For some regionals, the opportunity to discard select documents may well be what allows them to stay in the program. Imperfect though it may be, the Regional Discard Policy strikes me as by far the lesser of two evils.

  3. James A Jacobs on Mon, 1st Aug 2016 2:09 pm
  4. Hallie,

    Thanks for your comment.

    Perhaps we were not clear enough in the above post. The main point we intended to make was that the policy will result in loss of information. Is that the “lesser of two evils”? We believe that the choice is not between staying in the FDLP or leaving the FDLP. It is between preserving information and losing information.

    That said, you are, of course, correct that nothing *prevents* libraries from relinquishing their depository status entirely at any time. But the Discard Policy is not about permitting something that already exists. It is a blatant change in direction for the FDLP.

    The FDLP Regional status is instantiated in the US Code and the commitment was always seen as long-term. Libraries used to take this seriously as a commitment to public service and preservation. Now, as you point out, some are essentially saying “We won’t stay in the program at all unless you let us renege on that promise.”

    The new policy turns the FDLP commitment upside down. Where once the commitment was explicitly long-term, the new policy is short-term by design. Where once the commitment was to the public good, the new policy is *designed* only for the benefit of the discarding library.

    We now know that GPO’s implementation of the policy will not correct the shortcomings of the policy, it will amplify them and increase the risk of loss from a possibility to a certainty.

    We are also concerned that the very existence of the policy will *encourage* discarding where the old policy *prevented* it. Look at Regional Depository Library Discard Intent Survey. You will notice that, when GPO asked regionals which of 60 titles they would discard if they could, most libraries said they wished to retain 50 of those titles. We worry that, because the policy will turn the commitment upside down and make it easy to discard, it will *encourage* discarding by libraries that, in the past, did not even want to discard.

    We believe that the first and most important criterion for evaluating the effects of the policy and its implementation is asking, “Will this benefit the users of government information?” If the answer is “no”, then we should ask how it can be fixed.

    James and Jim

  5. Arlene Weible on Tue, 2nd Aug 2016 12:21 pm
  6. James and Jim: While I certainly support your ongoing efforts to raise questions about policy shifts in the FDLP, I find your overwhelming confidence in the current FDLP Regional structure as a sufficient model for long term preservation of government information baffling. There are no guarantees in the current system that Regional libraries will proactively preserve their collections … they only agree to keep what they receive on a shelf and hopefully catalog it. GPO really has no enforcement power when it comes to Regionals preserving documents in the current system. While I agree there needs to be some improvements to the current draft of the Preservation Steward MOA, I think it is still better to have “temporary agreements” with specific responsibilities outlined rather than rely on an old statute that doesn’t have anything specific to say about preservation practices. I also think assuming that libraries will jump out of temporary agreements at the first opportunity is somewhat disrespectful … I think most institutions do their best to honor the agreements they enter into. The long term prospects for preserving FDLP documents are better with Preservation Stewards than many of the current Regionals in our system because expectations will be more clear. Pursuing these initiatives are not ruining a great system, it is adapting a program to the real environment we live in.

    My two cents!

  7. James R. Jacobs on Tue, 2nd Aug 2016 5:05 pm
  8. Arlene,

    Thank you for your comment.

    It seems the substance of your comment is that FDLP could use better preservation standards than currently exist (we agree!) and that the Discard Policy’s “specific responsibilities” are better than Title 44 (we disagree strongly!).

    Since you did not say which responsibilities of the new policy you think are better than the existing ones (44 USC §1912, and the Legal Requirements & Program Regulations of the Federal Depository Library Program) or which “expectations” are “more clear,” we can only respond by repeating why we think the new policy will be much worse than the good (even if imperfect) system it replaces.

    The Discard Policy and the implementation as documented by GPO will result in the loss of information in two ways.

    First, the Policy allows good paper copies to be discarded but provides inadequate procedures for ensuring that the few remaining copies will be complete or undamaged. Need we point out that there are not hundreds or even dozens of copies of these documents floating around in used bookstores or elsewhere? There are certainly few if any complete collections of important series outside of the FDLP. When we throw them away they will be gone. If we make a mistake because of a badly implemented, badly designed policy, the loss will be irrevocable. Do the math: a policy that relies on 4 copies must be better than ten times more reliable than a policy that relies on 47 copies. The current system is not infallible, but it can fail repeatedly without losing information. The regional discard- and concomitant “preservation steward” policies cannot.

    Second, the Policy relies on the assumption that users do not need paper copies because they will only rely on the digital copies. If that assumption is false, the premise of the Policy is fatally flawed. And we know it to be false because of existing research (Conway). If 35% of those digital copies cannot be used as reliable surrogates, what will users do when they need those books? Will they travel to one of four libraries to get a copy and wait for it be brought to a “special collections reading room” (and remember, the “preservation steward” idea assumes no ILL!)? If they need two books, will they have to go to two different “stewards”? Is this a service model that you really think is “better” than the current system? Better for whom?

    Note that, over the last two years we have repeatedly called for additional research that would inform a discard policy to ensure against information loss, but GPO (and supporters of discarding) have ignored those calls. The result that we have before us is not encouraging.

    We are not saying that the current system is perfect. Neither are we saying that it cannot be improved. We are saying that the new Policy increases the risk of loss from a possibility to a certainty.

    We are glad that you agree that the MOA can be improved. We would welcome your suggestions and hope that you will share them with GPO and the FDLP community. We also suggest that it is not just the MOA that needs improvement. We would hope that GPO would provide a comprehensive document that spells out the requirements, standards, and procedures more thoughtfully and consistently than these separate drafts do.

    While on the topic of improvements, let us spell out a few that we would like. We would like to see a system where regional responsibilities for preservation are better spelled out and outlined, a system in which GPO assists with curation and cataloging, a system in which GPO assists regionals in alleviating their particular space issues (letting a little air out of the system in strategic spots) instead of scrapping the entire thing because a few regionals need “flexibility.” We think this would be better than what we have and what the Policy offers. We think it would be much better than a system in which “preservation stewards” sign MOAs that can be cancelled at the whim of the next library director. (This is not disrespectful, the Policy and the MOAs shift the balance from a system supported by the US Code to one where library directors get to decide based on local needs rather than national interests.)

    We would like to see GPO and FDLP design a system that looks forward, not backward, in digitizing the Historic Collections, a system that emphasizes access and long-term preservation by addressing the current and future needs of users of government information. We have outlined a proposal that, we think, would do much of that. We welcome your comments or alternative proposals.

    This policy has had a single explicit goal all along: discarding paper copy documents. No one has ever successfully argued that the policy is necessary to help users or that it will. We believe we have shown that, rather than help users, the policy will actually be detrimental to the needs of users. Most of those who support the policy make vague claims about “the real environment we live in” and “the future” but fail to address the real (actual, documented) problems of the policy or provide any real, substantive alternatives — and worse, attempt to trivialize the arguments of those who are pointing out that the emperor has no clothes. If the policy is bad for users, we should fix it, not make vague excuses for why it is the “lesser of two evils” or might be better than an unspecified alternative.

Tell me what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!

  • Our mission

    Free Government Information (FGI) is a place for initiating dialogue and building consensus among the various players (libraries, government agencies, non-profit organizations, researchers, journalists, etc.) who have a stake in the preservation of and perpetual free access to government information. FGI promotes free government information through collaboration, education, advocacy and research.
  • Archives

%d bloggers like this: