Strategic Planning Part III: Building a Collaborative FDLP

May 10, 2016 by
Filed under: Library, post 

In this part of our three-part series on strategic planning for the FDLP and GPO, we offer a vision of a collaborative FDLP that will greatly enhance preservation of government information, improve access for users, and increase the value of individual FDLP libraries to their communities and to the public.

Part I of this series suggested small changes to GPO’s strategic plan. Part II offered an informal SWOT analysis to give the strategic planning process some additional context.

Context: The Digital World and Digital Libraries

We believe that libraries have the same essential roles in the life-cycle of information in the digital age as they had in the paper-and-ink age. These include selecting, organizing, and preserving information and providing access to and services for that information. Libraries do this for specific communities because each community has its own specific information needs. Although the roles have not changed, the format of information and the methods libraries use to fulfill those roles have.

The digital world actually creates two big advantages for libraries. First is the ability to expand their user-base. For the most part, in the physical world, a library served communities who lived or worked near the library. In the digital world, this constraint no longer exists and a library can provide collections and services to communities literally anywhere in the world.

Second, digital objects in a digital collection are not constrained by physical limitations. The library is no longer constrained by the physical location of the digital collections it manages. A single document or file or database can be stored anywhere in the world and yet be part of many different collections and can be used by many people at the same time in many locations. This aids preservation too because copies of the same document or file or database can be stored and preserved in multiple locations while being part of one or many collections.

The fact that the digital world removes constraints that once limited what libraries could do and for whom they could do it (and how they did it) means that libraries can now add more value to information and they can reach more people with their collections and services than they ever could in the past. Libraries can expand their services, increase their service area, and serve more people better.

The two problems that keep this from happening are copyright restrictions (which re-constrain the use and sharing of digital objects) and the attitudes of libraries that self-limit who they serve (thus re-constraining libraries to serving a local-institutional community). Neither of these problems need constrain libraries from collaborating to provide access to government information. Digital government information that is in the public domain can be shared without the constraints of physical objects or copyright.

The Concept: A Collaborative FDLP

FDLP libraries, facilitated by GPO, can use this digital world to build a collaborative FDLP that is bigger, better, and more secure than GPO — or any one library — could build on its own.

Freeing each library from the constraints of delivering information only to a geographically-limited and institutionally-based community works two ways. One library can deliver collections and services to communities everywhere and every institution can receive collections and services from anywhere.

Thus, FDLP libraries can work together to provide, collectively, more than GPO — or any one library — can provide on its own. A collaborative FDLP is not one mega-library with one huge collection of only those documents that GPO can get. A collaborative FDLP consists of many curated collections that include Title 44 content, fugitive content (which GPO cannot force agencies to deposit), and non-Title-44 content that is out of GPO’s scope (e.g., FOIA’d documents, state/local/international government information, non-government information etc.). And each curated collection will have accompanying services tailored to that content for a community of users.

In such a collective approach, every community has access to the content and services it needs and every library provides a small slice of all those customized collections and services. In this approach, each library’s local-institutional community benefits from the contributions of every library.

This approach requires libraries to make one big change in the way they think of "communities." In this approach, a "community" is a group of people who have common information needs — they need not live and work near any particular library or even near each other. In this approach every library focuses on one or more Designated Communities.1 In this approach every institution benefits from the collective work of all FDLP libraries rather than the individual work of only its own local-institutional library.

This approach will result in an FDLP collection that is more complete than GPO can build and maintain on its own and more comprehensive than Title 44; it will have much better functionality, and it will be more secure for the long-term.

Digital Collections

How would this work? FDLP libraries would house and control digital collections. Any FDLP library could use one or more of three different ways to build digital collections:

  • Actual. Libraries will be able to acquire copies of each digital object they wish to add to their collection, store (i.e., "house") that copy and preserve it. They will also provide access to it to libraries and individuals through digital services, metadata, and APIs.

  • Virtual. Libraries will be able to use digital services, metadata and APIs to build "virtual" collections of items housed and maintained by others. The library will control the integrity, location, and permanence of the items in such a virtual collection though service agreements (formal and informal).

  • Hybrid. Libraries will be able to build collections using both actual and virtual copies of digital objects.


FDLP libraries will be able to select content in many ways. In all cases, the primary selection criteria would be based on the needs of a specific Designated Community. That could be as general as the "General Public" or as specific as "K-12 students of Civics." Examples of selection methods include:

  • Provenance. Some libraries will choose to focus on acquiring digital copies of all the information output of one or more agencies (or sub-agencies). This method addresses the needs of communities that need access to content based on provenance. Advantages: Guarantees comprehensiveness; keeps the provenance of the content in context; makes tracking versions and editions easier; is relatively easy to automate acquisition through the use of web-crawlers and similar technologies.

  • Subject/Discipline. Libraries seeking to build strong collections that address the needs of a subject or discipline or other category of information, regardless of the agencies that produce that information, will select individual content based on those criteria. This kind of selection might include selection based on (for example) a person (the Congressional representatives of a State), place (all documents pertaining to a state or region), kind (only agricultural maps or only geological maps), or time (documents of a particular event or period). Advantages: Focused on very specific community needs; frees content from agency "silos"; enhances browsing and searching; user-focused rather than producer-focused; same content may be selected for different collections thus enhancing preservation and access.

  • Format/Utility/Functionality. Some libraries will select digital content based on the format or utility or functionality of that content in order to address the needs of users for that functionality. For example, a library may choose to select only PDFs for formatting and human-readability where another may select formats that are more suitable for computer analysis, reformatting, and reuse. Such selection might also include selecting social media posts, videos, multi-media, data, GIS, etc. Advantages: Aids discovery; allows for customized presentation and delivery of content; allows for specialized functional services.

Advantages of a Collaborative FDLP

A collaborative FDLP will benefit the communities served by FDLP libraries and, by doing so, will benefit the FDLP libraries and their parent institutions. Here are some of those benefits:

  • Library Efficiency. One of the most important advantages is that a collaborative FDLP reduces the need for every library to provide the same collections and services as every other library. As long as a collection and service is provided by one library, it is available to all — thus freeing every other library to create their own collections and services for sharing. This means that a library that can only afford to build one collection and service can do so, but can receive the benefits of many collections and services. Similarly, two or more libraries can collaborate to build a single collection. For example, a library of Congressional Hearings might be built by many libraries with each focusing on only one committee.

  • Document Efficiency. A single comprehensive preservation collection of documents from a single agency could be used to build many different virtual collections in many different libraries for many different communities. Content from many different sources (and stored in many different places) can be assembled into many different kinds of collections. For example, a collection based on a state’s Senators could include their official documents and reports and statements on the floor of the Senate as well as their tweets and other social media and interviews and transcripts and videos from non-government sources. A single document from NASA might be in an agricultural collection, a space-science collection, and an art collection. Several libraries might select and store and preserve the same document for different reasons for different communities. In practice, a digital file stored by one library may be part of collections at many libraries. Those libraries can build collections that meet the needs of communities that may be local or far away.

  • Effective Preservation. Several libraries may store copies of the same file, thus providing redundancy of storage and enhanced protection against loss or corruption of the file. Files that are selected by more than one library (for different reasons, for different communities, for different collections) will be more likely to be preserved (and used) than files preserved only once and hidden in a single-mega collection. Since even virtual libraries will have control over collections housed in actual digital libraries, content will be less likely to be deleted or lost or altered due to carelessness or waning interest or loss of budget.

  • Uninterrupted Access. When there are several copies of a document or file or database, a collection can point to all those copies (using DOI or similar technologies) thus reducing the problem of loss of access for technical, political, economic, or other reasons.

  • Discovery, Utility, Service. By developing services that are closely tied to a collection, it will be possible for libraries to better address the needs of the target community. Small collections of the information needed by a community are easier to search and browse than mega-collections. Collections that are organized and delivered based on the needs of a community of users are easier to use than collections based on producer or agency. Different communities need different kinds of functionality from digital objects. By ensuring that a community can more easily find the information with the right functionality, libraries add value to the content for the communities.

  • No more silos. No longer will users of government information be forced to understand the government agency hierarchy or search in many different silos of information. A collaborative FDLP will offer focused collections built around the way the Designated Community looks for and uses information. A collaborative FDLP is not a thousand silos with information hidden and hard to find; it is one library with user-focused access-points (collections).

  • Many levels of participation. Even libraries that do not have the budgets or staffs to build digital infrastructure can participate in a variety of ways. For example, tracking the publication output of an agency; reporting "fugitives"; creating and enhancing metadata; reporting inaccurate and illegible digitizations; tagging content within documents (images, tabular data, maps etc); liaising with other library communities; providing training and instructing for users, librarians and LIS students.

Advantages to the larger library community

In addition to benefitting the FDLP community, a collaborative FDLP will also benefit the larger community of libraries including non-FDLP libraries, the communities they serve, and their parent institutions and agencies.

The first-order effect of a collaborative FDLP will be that it provides collections and services to everyone, including individuals (whether or not they are affiliated with a library) and non-FDLP libraries and the communities they serve. This benefits individuals and those libraries that have communities that need government information.

A byproduct of that effect is that non-FDLP libraries will get the benefit of better access to government information without having to expend resources. That will free resources for use on other projects. The communities those non-FDLP libraries serve will get better access to government information and they will get better services as their local library allocates its limited resources to non-government information projects.

Another byproduct of the collaborative FDLP is that it can serve as a model for other collaborative library projects. There is a wealth of information that is not burdened by copyright that could be made much more discoverable, preservable, and usable in collaborative projects much like FDLP. The knowledge base, skills, infrastructure, training, and formal collaboration agreements can all be used either directly, indirectly, or as models for similar projects.


What is preventing us from creating a collaborative FDLP ecosystem? Some would say that the biggest roadblock is lack of resources. "We don’t have the budget. We don’t have a data center. We don’t have the staff." These are certainly real issues, but they are not insurmountable. In fact, "lack of resources" is more of an excuse than it is a roadblock. Our parent organizations are unlikely to fund us to do something that we have never done unless we make the case for doing it first. We address how to do this below.

What about "free-riders?" A "free-rider" is a library that receives the benefits of a public good without contributing to the costs. There is, of course, the risk that some library administrators at FDLP libraries will conclude that their library and their local community can get all the benefits of all those nicely curated collections and all those specialized services without participating in the FDLP, without providing any collections or services themselves, without expending any resources. There is a risk, therefore, that FDLP libraries will drop out of FDLP and put their resources into other projects. We believe that an effective collaborative FDLP will actually do the opposite of this; it will reduce the temptation to leave the FDLP and provide strong incentives for others to join. We address this below.

Building the case for a Collaborative FDLP

Building a collaborative FDLP is going to require convincing library administrators and parent institutions to participate. These are some of the reasons that participation will be attractive to individual libraries:

  • As noted above, participants will get more for each dollar they invest in collaboration than they would get by investing that dollar in isolation.

  • Collaboration will work if libraries participate in good faith and will fail if libraries attempt to "free-ride."

  • Failure to collaborate will result in loss of information because neither GPO alone, nor all government agencies can or will be able to preserve their own information. This is a simple fact because of the limitations of the missions and legislative mandates of those agencies, and because of the bureaucratic constraints on those agencies, and because of passive political inertia and active political opposition.

  • Collaboration will enhance the value of the library to its own community. Collaborators will be able to actively participate in shaping the collections and services that are available. Collaboration will ensure that the library’s own local-institutional community will get more and better long-term access to government information than the library could provide on its own, or that GPO could provide without a collaborative FDLP, or that the unpredictable ebb and flow of the Internet would provide.

  • Collaboration will enhance the value of the library in absolute terms. As libraries move away from using simple metrics of value such as collection size, they will increasingly use more subtle — but more substantive — metrics such as participation in collaborative projects that demonstrably enhance preservation and access, and development of new kinds of resources including expert digital library staff, digital library services, and digital library collections.

  • Collaboration will enhance the value of the library in comparison to non-FDLP libraries. FDLP libraries that build digital collections (whether actual, virtual, or both) will control those collections in a way that non-FDLP libraries will not. This means that participants will decide what information is selected for preservation, how it is presented and served, and how it is delivered.

  • Collaboration will guarantee the preservation and long-term free access to government information. This information has inherent value. Collaboration provides libraries the most control over what will be preserved for long-term free access. All other options will result in either unpredictable loss of valuable information or privatization of that information.

  • Participating FDLP libraries will be able to develop large-scale proof-of-concept digital library services that would be more difficult or more expensive to develop for copyrighted content. Such projects will add value to library services. This will attract increased funding and other support. It will also serve as a model and offer tools for other digital library projects with non-government information.

  • A collaborative FDLP will demonstrate to the General Public the value of libraries working for the common good and will contrast that value to the options of relying on commercial services that either charge fees or invade privacy to provide access to information. Libraries that participate will be viewed as leaders with a vision.

Next Steps

How can GPO and FDLP and individual libraries and librarians begin the transformation to a collaborative FDLP? Here are some suggested first steps.

  • Web harvesting. If agencies conformed to some simple standards of web site design, it would be easier for GPO and others to crawl and harvest those sites. This small change in procedures would benefit agencies by knowing that GPO and others would take the burden of preserving their web-based information. These procedures could include simple web-site structural details such as including information in standardized directories (e.g., and The Office of Management and Budget could establish and enforce such simple standards. Action: Develop a proposal for OMB, gather support from relevant groups (e.g., Open The Government, OpenHouse, ALA, AALL), submit and lobby for it.

  • Metadata. Similar simple standards for incorporating simple metadata (such as Dublin Core) into web-based information would facilitate the re-use of metadata and building a National Catalog of government information, indexing web harvests, and versioning documents. Action: Develop a proposal for OMB, gather support from relevant groups (e.g., Open The Government, OpenHouse, ALA, AALL), submit and lobby for it.

  • Preservation of FDsys/ Although, GPO has FDsys content stored in two locations, they are both within 100 miles Washington DC. Although GPO has graciously facilitated the LOCKSS-USDOCS program of collecting and preserving all FDsys collections, there is no official partnership with LOCKSS-USDOCS. In order to enhance preservation of information in GPO’s collection and facilitate GPO’s compliance with OAIS preservation standards, LOCKSS-USDOCS and GPO should enter into a formal agreement as preservation partners. Action: lobby GPO and LOCKSS to enter into an official partnership.

  • FDsys/ alternatives. Currently, LOCKSS-USDOCS is a preservation system only and provides no access to the content until a “trigger event” occurs (system or organizational failure, catastrophic event etc). In the event of loss of access to GPO-hosted content, it would take some time to make the LOCKSS copies available. LOCKSS should work with its partners to develop public interface(s) to all or selected parts of the collection. Projects could include providing simple access to all LOCKSS-USDOCS data; developing access interfaces that address the needs of specific communities; developing APIs for sharing individual documents and metadata-defined groups of documents. Action: Open discussions with LOCKSS and its partners.

  • Funding. Model funding proposals by individual FDLP libraries and groups of FDLP libraries working together on joint projects for preservation, metadata creation, and service delivery could expand the possibilities for the collaborative FDLP. Such model proposals would articulate the benefits of the collaborative approach and could serve as models for other projects. They could be aimed at institutions, library consortia, and grants organizations. They could target information that is not in FDsys. Action: Establish working groups within FDLP, GODORT, ARL, AALL, and other organizations.

  • Universal Registry/Catalog. There is no complete union catalog of printed government documents and no complete national bibliography of digital government documents. A collaborative FDLP will require both. There is a need to coordinate and integrate and enhance existing projects (e.g., The Catalog of Government Publications (CGP) the registry of digitization projects, the HathiTrust Registry of Government Documents). Action: Create a working group to chart a path forward to build a national bibliography and union catalog.

  • Coordination. Once basic parameters for a collaborative FDLP are set out, collaboration can emerge formally and informally. Every new project will contribute even without elaborate or formal Service Agreements. Nevertheless, it will be useful to have a registry of interest and wish-lists, project-planning, and project completions. Action: Building on the ideas and work of ASERL’s "Centers of Excellence" create an online collaboration tool to track progress.


It was almost twenty years ago (June 1996) that the era of the so-called digital Federal Depository Library Program was initiated. And yet, neither GPO, nor the FDLP as a whole, nor FDLP member libraries have any of the benefits and guarantees of being “depositories” or “libraries” or a “program.” Even worse, the communities that FDLP libraries serve have none of the benefits that a local depository library once provided.

After FDLP libraries allowed (and in some cases even encouraged) GPO to intentionally replace a many-institution, decentralized model of preservation of and access to government information with a single-institution, centralized model, we can see clearly that the new model is not just inadequate, but dangerously fragile. Most government information is not being preserved by GPO and federal agencies have neither the mandate nor the will to do so themselves. Access is through agency-focused silos rather than user-focused collections + services. Libraries have no control or voice over decisions to alter or move or remove documents or entire websites. Such losses often leave libraries with either no access to needed information or only with commercial fee-based access.

Given this state of affairs, libraries are faced with two alternatives: do nothing and lose information or take action to ensure long-term preservation and access that GPO cannot guarantee on its own. We suggest that doing-nothing is simply unacceptable because of the inherent value of government information.

We believe that the best option for taking-action is for individual FDLP libraries to work with GPO and other FDLP libraries to create a Collaborative Digital FDLP. To do this, libraries need to participate in the life-cycle of government information and actively create value for communities that need government information. A collaborative FDLP will provide better infrastructure, collections and services to all users of government information and will be a model for other libraries and collaborative projects. Let’s make it so, together!


Appendix I: The Collaborative FDLP in Practice

The heart of the collaborative FDLP is that it relies on libraries sharing responsibilities. Each FDLP library must be assured that its own, local-institutional communities have the collections and services they need, but these collections and services may be provided by one or more FDLP libraries. The collections and services provided by any FDLP library are open for use by anyone.


  • Any FDLP library can host one or more collections or services.
  • Any FDLP library can house one or more collections.
  • Each collection and each service would be aimed to serve a specific Designated Community.
  • Designated Communities need not be geographically-based. Members of a DC can live or work in one location or many locations or in no specific location at all.
  • Every collection and every service will be available to all communities — including non-FDLP libraries and individuals.
  • A master register of individual contents, collections, and services will serve as a Union List, will tie all collections and services into one master collection, and will provide an additional, overall entry point to the FDLP.

Appendix II: Definitions

  • Collection is any group of documents or files in any format along with the metadata that describe the individual items in the collection and the collection as a whole. A collection need not be housed in a single location or with a single library. It need not be housed with the hosting library. Each collection is controlled by a host.

  • Control of a collection. A hosting FDLP library "controls" a collection by specifying its contents. No materials may be added to or deleted from the collection without the approval of the host. The host also controls the metadata defining the items in the collection and the collection itself. Both the metadata and the items in the collection are completely open for use by others.

  • Designated Community. An identified group of potential consumers who should be able to understand a particular set of information. DCs can be very broad (e.g., the General Public, "undergraduates," etc.), or very specific (e.g., "graduate students working with historical demographic data"). DCs can also be defined as users of specified content (e.g., "users of Congressional information" or "users of FOIA documents").

  • Hosting a collection means that the library has virtual control over the contents of the collection whether the library houses the collection or not.

  • Housing a collection means that the collection is under the physical control of the library.

  • Local-institutional Communities are those communities that are part of a single institution that funds a single local library. Colleges and universities and schools have communities of students and researchers and fund their college and university and school libraries. Cities and counties and states have communities of residents and citizens and fund their public libraries. Government agencies have communities of their employees and fund their own libraries.

  • Services include computer-mediated access to the collection (software, UI, UX, APIs, etc.), databases of metadata, and human reference services.

  • Service Agreements. Formal service agreements between FDLP libraries regulate the control of collections and their hosting and housing. A library that houses a collection my provide a "Public" service agreement that defines the collection and specifies the services that it will provide to all.


  1. “Designated Community” is an essential concept of OAIS. It is defined as “An identified group of potential consumers who should be able to understand a particular set of information. The Designated Community may be composed of multiple user communities. A Designated Community is defined by the Archive and this definition may change over time.” Consultative Committee for Space Data Systems. (2012). Reference Model for an Open Archival Information System (OAIS) (Magenta Book, issue 2). Washington, D.C.: Consultative Committee for Space Data Systems.  ↩


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: