An interesting and quite broad collection of public and school library policies on several subjects may be found at Sample Library Policy Statements.
Public Library Policies and Plans Online by WebJunction includes a good collection of policies and practices.
Perry Willett has created a directory for policies that deal with digital resources.
The Collection
Development Manual of the National Library of Medicine is a
good example for a comprehensive research collection.
Selection is the information professional's primary
responsibility. All other responsibilities are related to
this one. Thus, the creation of the collection development and
management policy is the cornerstone. The policy should
provide "an intellectual or conceptual framework for rational
decision-making." The policy, or supporting documents, must be specific
enough to allow consistent application. The policy must change as
situations change so it cannot become a static document.
The purpose of the collection development policy is to determine and guide the present and future state of the collections. The policy should answer the questions:
Ideally, the policy statement would include an accurate characterization of past collecting, of current collecting, and goals for future collecting. The policy is designed to insure that collection development and management is focused and systematic while being clearly linked to the success of the parent institution. The use of collection levels for particular subject chunks is the heart of the process because that is where the priorities are most clearly evident.
There is no common, uniform or standard definition for a
collection policy statement. The literature suggests that there are as
many models as there are selectors and collections. Most policies are
vague and do not contain clear, persuasive priorities. Since
collections contain many different objects, some tangible [a collection
of goats] and some less so [digital images], policies will vary in
particulars. Still, all should contain the major elements seen below.
While not covered here, there is a rich tradition of policies for
scientific collections, including living ones.
In smaller information agencies, with smaller collections, there may be one over-all policy. In larger, research-oriented agencies, where collections are larger, there are likely to be several policies. These policies are likely to deal with four topics:
A policy should deal with an important and continuing problem. The policy saves time and effort because staff know what to do when faced with a particular problem. Without a policy, a new decision must be made each time a problem appears. This decision making from the beginning and without formal guidance is sometimes called ad hoc ism. If you face a collection development or management problem once every three or four years, you probably don't need a policy. If appraisal of gift books is a problem once or twice each week, you do need a policy.
The CDM policy (or policies) should be useful and should be used. Problems should be solved quickly and with minimal (considering the circumstances) time and effort. For the collection developer, the heart of any policy statement is its list of priorities . The policy statement should contain a persuasive and orderly expression of priorities useful for developing and managing the collection.
The CDM policy may be written or explicit or formal. It may also be unwritten or implicit or informal. An effective policy must be followed by staff and many information agencies have unwritten policies that are understood and followed. Never assume that lack of a written policy means that there is no policy. In fact, some organizations have more unwritten policies than written ones. Often, unwritten policies accumulate haphazardly over the years and may be difficult for the new hire to identify and understand. If you are new to an organization, it is important to ask colleagues about unwritten policies.
There is a strong relationship between collection size and the availability of written policy statements. In general, few collections have a current, persuasive, and organized policy statement. Large collections are more likely to have written policies than smaller ones.
There is also a relationship between controversial material and policy focus. Collections that might contain controversial material which could invite censorship incidents are likely to have policies that focus on complaint-handling. These policies are also more likely to provide a more detailed rationale for collecting subjects that might be controversial.
Collection policies should reflect the agency and parent institution mission as well as appropriate supporting goals and objectives. One of the best rationales for collection development is that it helps the parent agency or organization to be more successful. Linking collections to organization or community success is essential. For example, the University of Tennessee, Knoxville has a goal of attracting and retaining African-American students and faculty. As information professionals, we need to understand how library collections can help the University to attain this goal.
As an aside, collections may also have associated objectives or specific, measurable targets or outcomes. Although not frequently encountered, such targets would answer these questions:
For example, you may decide that by a certain date the collection would hold all of periodicals found in Business Periodicals Abstracts. The cost could by found via a periodical vendor. The outcome would be a certain desired level of access to the core of business periodicals.
A good policy deals with an important, continuing problem. It is based upon a thoughtful analysis of parent organization and information agency mission, goals, and objectives. It considers what will be emphasized in the collection and what will not be emphasized as part of a clear focus on audience. It will certainly consider available resources, including personnel. It is clearly stated in as few words as possible using standard English language with minimal professional jargon. The policy is reasonably stable--that is, it should serve for three years or so without major change. A good policy has been adopted by the appropriate governing authority so that it has standing and can be enforced. A good policy is one that has been explained to those who will implement it. These people have also had an opportunity to participate in creating the policy. A good policy is used.
A written collection policy has several assets. It should be more easily understood since it can be read and reread as needed. This should encourage self-examination and reflection. Written policies are easier to distribute or exchange. They can be more easily analyzed or interpreted in training or orientation situations. Consistency of action is more likely with a written policy since it can overcome both time and space. The written policy is also less likely to be misunderstood. Because it is more formal and is often adopted, the written policy is more likely to prevent inappropriate action and may protect the collection against undue community pressure. If the information agency is participating in some sort of cooperative initiative, the written policy promotes cooperation and coordination by making priorities, policies, and procedures more visible. Similarly, the written policy may help to support resource allocation decisions by indicating that selection is a thoughtful, quality controlled process guided by appropriate priorities. A written policy may provide explicit recognition of the agency's commitment to intellectual freedom.
As a statement of priorities and desired action, the written policy may also be a planning document. It can also be used as a yardstick for evaluation. If policies are not controversial, the written document should strengthen public relations by sharing the good news that collections are developed and managed systematically and rationally to meet genuine community needs. When shared, the written policy makes the collection more accountable to the community.
Any written policy has certain liabilities. It is more costly in time and effort, particularly because of the adoption process. Often, this creates a situation in which adopted policy is not revised because it is too much trouble; thus, encouraging the status quo. Written policies are more likely to be dated. Written policies are less flexible than unwritten ones and may discourage initiative and creativity. Because they are more easily shared, written policies are more likely to encourage criticism and sniping by community members and politicians.
In this course, the primary focus is on collection development policies in libraries. However, many other agencies, especially non-profit ones, have collection development policies. For example, the Society for the Preservation of Natural History Collections has an excellent and somewhat lengthy set of guidelines for natural history collections. For example, their first objective is:
"Management and care of collections of natural history materials should be governed by respect for the scientific, historic, physical, cultural, and aesthetic integrity of the specimen or artifact and its associated data. Concern for the future should include protection against unnecessary damage, loss, or alteration that might affect its future research, educational, or exhibition potential."
An example of one of their responsibility statements:
"Each institution should develop collections policies and procedures that provide a written framework for collection management, care, and use. It is essential that each institution also provide the resources (e.g., time, money, qualified personnel, appropriate space, and facilities) needed for the long-term preservation and documentation of the collection under its responsibility, or make alternative arrangements for collection management and care with an appropriate allied institution."
An example of one of their staff responsibility statements:
"Collection care personnel should have appropriate training to understand fully all aspects of collection work (e.g., legal, ethical, environmental conditions, management, security, health and safety), the limitations of their own expertise and authority, and the consequences of any decisions and/or actions they make take or recommend. Every effort must be made to consult with appropriate specialists to ensure that all aspects of management, preservation, and use are considered before authorization for actions is given."
An example of one of their use of collections statements:
"Use of collections should be carried out in ways that are compatible with preservation objectives and concerns held by indigenous peoples, whenever possible. Certain specimens or artifacts may be considered too rare, fragile, culturally sensitive, or significant for exhibition or loan (e.g. type specimens, specimens of extinct species, historically significant specimens, or specimens in poor condition."
This is the first step. It may be the only step. We begin with the questions: "Is a policy needed"? "Why"? Without persuasive responses here, the process stops. Since policies affect action on repeated problems, there should soon be another opportunity to consider the matter.
Some information professionals, usually in large, complex organizations, believe in the "comprehensive approach" which advocates a policy with supporting procedures for nearly everything. Others favor the "selective approach" where policies are created only when the need is unavoidable. While I actually enjoy creating policies, the selective approach is best. This means that you might have a written policy for only one or two troublesome subjects, say sex education or creationism, rather than one for all subjects collected.
A few standard variables are usually considered in making this decision:
If a policy seems ad visible, a committee is usually formed to gather data, discuss, and create drafts. Those who will use the policy or interpret it to users should be represented. In some situations, it will be helpful to also include community representatives.
Policies need not be developed from scratch. Review of peer institution policies is usually helpful and may provide all the elements needed to create a draft document. The draft should be well-organized and clearly written. Drafts are shared with those who need to know and those involved in the approval process until the product is in good shape.
Draft-sharing is an important part of the approval process since it provides an opportunity to educate the approvers while learning more about their views and preferences. The key element here is "no surprises." Governing boards don't like to be surprised by a policy that contains controversial elements. Drafters don't like to be surprised by board opposition or misunderstanding.
Without adoption, the policy lacks standing in the community. With adoption, the policy is now a community one (community speaking through its representatives) and not simply an information professional decision. As you might expect, policy adoption often involves politics. Sometimes, compromise may be necessary. Other times, it may take years before the board can be persuaded to adopt a policy or a certain course of action. Lobbying may be necessary, but it needs to be done with tact and diplomacy. It is best done by community members representing themselves. Putting pressure on board members can result in setbacks.
Sometimes this step is called policy promulgation or policy dissemination. The key question is the degree to which the policy should be shared with those inside the agency and the community. It may be useful to begin with "Who needs to know?" and go from there. Variables to be considered include:
Since rumors and hearsay about new policies can create misunderstanding and anxiety, it is important to issue the policy as promptly and fully as possible. When policies are distributed, it is also important to collect and destroy super ceded policies.
Policy education insures that the policy is clearly understood. It begins by involving staff in the policy creation process. After the policy is adopted, meetings are held with ample feedback opportunity. Questions are answered and policy elements are explained until the policy is clearly understood.
Policy acceptance covers the range of activities needed to insure that the policy is properly implemented and complied with. Acceptance begins by emphasizing the benefits of the policy. Moral suasion is often used to encourage compliance because it is the "right" thing to do. Staff need to understand that compliance will be monitored and that senior management is committed to compliance. Finally, acceptance may require some indication that discipline will follow if the policy is not followed. Obviously, a document does not produce results. Policies may be ignored. It is the job of management to insure that does not happen.
Feedback is gathered on how well the policy works. Normally, there will be some confusion about some aspect of a policy so that it will need to be interpreted. These interpretations need to be shared with others and incorporated into a revised policy. Otherwise, there will be different policies by those who have shared in the interpretation and those who have not. Often, it is the interpretation which really defines the policy and makes it work.
This is the final step in the process. The policy is examined to see how well it has worked. Again, we ask "how often was it used?" and "Who used it?" Based on responses to these questions, the policy may be deleted or retained. If retained, we might return to the policy creation stage and begin again.
The policy format will vary with the length and complexity of the policy and the audience. Again, we could ask the questions mentioned above:
Thus, the policy could range from a few short paragraphs to two or three loose-leaf notebooks. Today, many collection policies will be in folders on information agency servers. Since there are few "typical" collections, there are only a few typical collection development/management policy statements. Since libraries tend to borrow policy format and element from peer institutions, it is common to find similar CDM policies for particular types of libraries.
Often, policies are divided into general and specific sections. General policies would apply to a variety of subjects, formats, or audiences. Specific policies apply to particular subjects, formats, and audiences.
The general policies might begin with contextual information about the parent organization and the information agency. This would include:
General policies provide information on the environment in which CDM activity takes place and introduces activities likely to interest users or potential users.
Criteria are concrete rules or standards used in making selection decisions. They may be informal (in one's head) or formal(written down and perhaps even adopted). They may be easily knowable (select no item with a list price of more than $200.00) or only somewhat knowable (select only works issued by reputable publishers). Criteria receive considerable attention in the literature. Some authorities focus on general criteria which should apply regardless of subject, audience, or format. Others emphasize unique criteria that apply to one subject, audience, or format. Given the substantial variety available, it seems reasonable to focus on more general criteria. Here are some examples in brief list format. Please note that these lists are NOT ranked.
While each format may have important characteristics and attributes to consider, here are some general criteria that apply widely.
Costs may be direct or indirect. Direct costs are relatively easy to identify and evaluate. Indirect costs are those that are associated with use of a particular item. For example, use of a software package may require hardware/software not now available as well as designated space. Staff orientation/training is also part of indirect costs. Depending on one's orientation, consumables associated with workstation use, for example, may be direct or indirect costs. Although not normally considered, we could also consider here user costs in time, effort, and convenience.
Often, criteria is general and applies to several topics, formats, and audiences. For example, selecting works by "reputable publishers" is general. In some cases, criteria will be subject specific. For example, selecting only works with English measures for a cookery collection.
Criteria may be used for exclusion and there is some evidence that fearful selectors have used selection criteria to exclude potentially controversial works. Typically, exclusionary criteria are vague and leave much discretion to the selector. Here are some examples of criteria that have been used to exclude good works. Do not select an item with:
Nearly all selection criteria could provide a rational for refusing to select a controversial but worthwhile item. Remember that selection focuses on reasons for inclusion while censorship focuses on reasons for exclusion.
Most selection policies will contain some individual subject policy statements. These subjects are usually tied to specific ranges of a classification scheme. The policy format found in the separate process handout is fairly typical of what might be encountered in an academic research library. Some subjects may also include subject-specific criteria in addition to the general criteria mentioned above. For example, cookbooks may need to have exact measurements or use commonly available ingredients.
Typically, each topical listing would include five elements:
There may be a separate policy for each major format (microfiche) or one policy for similar formats (micro formats might be an example). It may also be appropriate to include format as part of the subject policy. However, that might require frequent repetition of common elements.
When there is an audience with unique needs or problems, it may be helpful to have policies for a particular audience. For example, public libraries usually will have a separate policy for selecting children's materials. Academic libraries may have a separate policy for managing collections to meet the needs of distance education students. Again, audience may be included under the subject policy.
The Research Libraries Group Conspectus was designed for
evaluation and resource-sharing purposes. However, it can also play an
important role in the subject breakdown and levels section of the
policy statement. The goal of the project was to produce an instrument
for evaluating collection depth and then creating a database that would
contain depth characterizations of many subject collections. Thus,
researchers and information professionals could quickly discover which
institutions hold the best (deepest) collections on a particular topic.
OCLC and the AMIGOS Bibliographic Council developed software to make it
much easier to compare holdings with peer institutions.
Since the project was developed by research libraries, the LC classification scheme was used for subject analysis. Based on LC class numbers, 3500 subject fields were created. WLN, now part of OCLC, has created a smaller list based on the DDC scheme. Collection depth may be assigned to each field. In a large research collection, this would involve considerable time and effort.
Each subject aspect is summarized with a collecting level tag which indicates how comprehensive or deep it is. Often, different aspects will have different level tags as collections vary. The report which characterizes strength for the several aspects within a larger subject is called a "conspectus." The conspectus captures "existing collection strengths" or ECS. When used in collection development policies, tags may also be assigned to "current collecting intensity" (CCI) and, most important, "desired collection intensity" (DCI). Sometimes, these terms are used instead:
The "0" tag is used to indicate an out of scope aspect where no formal collection development takes place. Information may be available within current periodicals or the reference collection, but no items devoted specifically to this topic will be selected.
The "1" tag is used to indicate a minimal collection containing a few basic, current works. Later, "1" was revised to add "1A" which minimal with uneven coverage and "1B" which is minimal, but with more balanced coverage, especially of standard works.
The "2" tag is used to indicate a collection containing basic information which introduces and defines the subject. Later, "2" was revised to add "2A" which includes basic introductory reference works, some retrospective material, general works on major topics, and major periodicals. In contrast, "2B" has a wider range of aspects and more depth.
The "3" tag is used to indicate an intermediate or instructional or study collection which will support undergraduate instruction and perhaps terminal MA/MS level instruction. Late, "3" was revised to add "3A" which focuses more on supporting undergraduate instruction. This collection contains a broad range of basic works in variety of formats, including classic retrospective items, major periodicals, reference and bibliographic tools. The "3B" collection includes the above, and more focus on the MA/MS course work and advanced independent study. This collection would contain more material on secondary topics, more retrospective items, and more emphasis on research techniques and findings.
The "4" tag is used to indicate a research collection which will support doctoral level research. The collection would be comprehensive with extensive back files of many serials in appropriate languages.
The "5" tag is used to indicate a comprehensive collection or an attempt to collect all notable work in all appropriate languages. This is an exhaustive special collection. Few information agencies would have level 5 collections.
The project also created language tags to summarize collection aspects:
You can see how the level and language tags are applied to library and information science by going to the UT Libraries Conspectus File for Library and Information Sciences. Note that level 3 is the highest found in these collections.
At one time the University of California campuses captured their collections on humanistic psychology as follows:
These might also serve as targets in a collection development policy IF the existing collection reflected appropriate priorities.
Subject breakdown and levels do not need to follow the model described above. You may create levels to meet your particular situation. For example, a collection on minor league baseball for a medium-sized public library might include these subject aspects.
There might be three levels with level one consisting of one or two basic works, level two consisting of three to five current general works that introduce and define the subject, and level three consisting of more than five items with a broader range of items including current and retrospective items and a variety of formats.
Depending on the community analysis then:
While designed for collection evaluation and resource-sharing, collection breakdown may also be used as a target in collection development (as done here). In some situations, you may find these tags used to capture:
Do tags measure collection depth in a reliable way? The evidence suggests not. For example, an examination of collections claiming level 3 depth in Russian history showed considerable variability. Using standard bibliographies for this subject, these collections held from 11% to 79% of standard works. It is difficult to believe that a collection with but 11% of standard items could support substantial undergraduate instruction. Without operational definitions, these tags seem to have little value.
More recently, evaluation has become more refined by
specifying which lists are to be checked and adding minimal holding
percentages for each level. For example, in psychology, a level 5
collection must have 60% of the titles in Psychological
Abstracts. However, UCLA has 69 percent and feels that it has
a level 4 collection.
Another problem is that the Conspectus measures holdings and
access may be more important today. Too, access to leased digital
resources may be ephemeral as access varies from year to year.
Select an information agency or an agency of your choice where gifts might be a problem. Consider creating a policy on "gifts." Identify audiences for such a policy. How would you go about creating this policy? Likely problems? Discuss how you would share this policy with the community.
You are a youth services librarian developing collections for a teen audience. Prepare a list of criteria appropriate for selecting novels. Which criteria are most important? Why? Can you create operational definitions for your criteria?
You have been asked by the Director of an academic unit to develop a collection of the best placement web sites for recent information science graduates. Prepare a list of criteria appropriate for selection. Which are most important? Why? Can you create operational definitions for your criteria?
Select an information agency of your choice. Select one of the "1" to "5" levels mentioned above. How might you make the level definition more operational and meaningful?
You are a new hire in a public library. The selection criterion used is "select works that provide insight into the human condition." How might you interpret this statement so that it provides some meaningful guidance?
