Database and Journal-Package Review 2017

The library is conducting a review of its subscribed databases and journal packages. The College's current initiatives to manage expenditures mean very rigorous examination of the Library's acquisitions spending. This review does not include books, DVDs, other single purchases, or subscriptions to individual journals; the latter we'll review another year.

See the resources under review, including links to title lists.

Our faculty/staff survey closed on April 12, and we are currently analyzing the results and combining them with our other criteria (see "How will you make the final cancellation decisions?" below) to develop a prioritized list of what we may be able to cancel with minimal impact on teaching and research at Middlebury.


About the Review

What is this project?

This is an assessment of the electronic databases and e-journal packages to which the library subscribes, in order to determine what’s currently most important and relevant to the College’s curriculum and research. Our general Collection Development Policy gives more information on our criteria for adding materials to the collection.

What’s the ultimate goal of this review?

The immediate outcome is to create a prioritized list of databases, so that when we get our actual FY18 budget numbers from the administration, we’ll be able to to act quickly to cancel the subscriptions necessary. The whole process takes time, and waiting until the Trustees meet in May is too late for us to gather faculty input before communicating our cancellation decisions to our vendors before the start of the fiscal year on July 1.

Why is it necessary?

For two reasons: First, regular review has been a normal procedure, in order to keep our collections focused most on supporting the curriculum and research, which are constantly changing and adapting to new knowledge. Second, the College is implementing cost-control measures that in many areas will require spending cuts. The cost of library materials rises far in excess of annual consumer inflation (click graph below for more detail), which means that even a budget equal to last year’s budget--that is, with no increase to cover that inflation--requires a reduction in the materials we can provide. Reducing our budget further means deeper cuts in what we can provide, and we want your input on those decisions.


How much do we have to cut?

The target will depend on the College’s final allocation for library acquisitions, which we won’t know until sometime in May. We’re preparing for a cut in funding of up to 3%, which has to be added to the average annual inflation in library materials--in recent years, that’s been about 5%. All told, if our acquisitions budget is cut by 3%, we’ll have to cut 8% of what we spent this year in order to stay within our budget for next year. That’s roughly $210,000.

Can’t we shift some funds from one-time purchases of books and DVDs to the subscriptions, so we won’t have to cut so much from databases?

Theoretically, yes, but that approach has serious ramifications. Because of inflation, the gap we would have to cover grows and compounds each year. It takes only a few years before our one-time funds available can no longer cover the inflation in subscriptions (see Chart 1). At that point, we would not only have zero funds to fulfill requests for books and films, but we would have to start cancelling databases and journals anyway. Even using one-time funds to cover only the first year’s inflation results in a loss that can never be regained (see Chart 2), unless the College restores the lost funding--plus all of the compounded inflation--to the library’s base budget sometime in the future. And even then, by that time we will already have refused a large number of purchase requests. Overall, the strategy looks good on paper for the first year, but the cost heavily outweighs the benefit.

Why do prices go up so much every year?

In part, these particular rising costs reflect the upheaval and uncertainty in publishing, especially commercial scholarly publishing. The market for scholarly information (i.e., higher education and research communities) is captive, in that there are some library resources we can’t do without, and publishers and other information providers know that. Their products are unique and targeted, and we need them in order to conduct our research and communicate its results to other scholars. In addition, consolidation of publishers and their parent companies over the last few years has reduced competition, also increasing costs. There’s a lot more to answering this question, but those are the major factors.

How are database prices set?

Database providers use many different pricing structures, the most common being based on the size of an institution’s enrollment.  Library prices for most periodicals are much higher than individual prices because one library subscription replaces a large number of individual subscriptions. Some companies also figure our prior year’s usage into their pricing, and some, unfortunately, seem to take endowments and tuition numbers into account as well. New content always raises prices, and publishers in particular are always buying and selling journals. Of course, our price from the publisher selling a given journal never goes down as much as the purchasing publisher’s price goes up.

Why can’t you get better prices?

We always try! Sometimes vendors are willing to negotiate, and sometimes we can go in with other colleges, sign multi-year deals, or get discounts for adding more than one new database or journal package at a time. However, some publishers simply refuse to negotiate, especially the large commercial ones who publish important journals that they know we need and can’t get anywhere else. Research materials simply cost a lot of money, and while we do find some success in keeping costs down, there are limits on how far our dollars can stretch.

How does the library decide how much to spend on each subject area?

While many college and university libraries allocate funds by department, we don’t. We seek out the materials that faculty, students, and staff request, regardless of discipline or department affiliation. The library is here to support all disciplines, and if in one year a particular subject area pretty much has what it needs, then there’s no problem using the funds elsewhere for the year. Some other year, a brand-new database, dataset, or other information resource might become available in that discipline, and we prefer to retain the flexibility to accommodate a request for that. Or, when a new faculty member with a new specialization joins us at Middlebury, we want to be able to meet those new needs. In the end, it all comes out of the same pot, and we try to treat everyone’s requests seriously and fairly.

Why don’t we spend the same amount on all disciplines?

This is a complex question with several different parts. Material costs vary widely from discipline to discipline, and different disciplines rely on different kinds of material. For instance, the sciences rely much more on journal articles than books, largely because things change much more rapidly in the sciences than in the humanities, and journal articles appear faster than books do. The concept of information lifecycle comes in here, but we’ll keep this as concise as possible. Scientific journal collections are extremely expensive, and the publishers most focused on scientific disciplines are those least willing to negotiate. Humanities databases cost much less, but we purchase many, many more books to support humanities disciplines than we do in the sciences, because books are still more important to the humanities. Social sciences are in the middle of that spectrum. Again, we try to focus on meeting everyone’s needs, rather than attempting strict numeric parity.

How did you decide what to review?

This review includes anything we pay for annually that provides significant library material, which means packages with more than a handful of journals, or databases costing more than about a thousand dollars. Some of our annual fees are lower, but they pay for maintaining access to and/or updating content that we purchased on a one-time basis in the past (that information appears in the notes on those specific database). Some subscriptions are staff resources for things like cataloging materials so you can find them. We have already cancelled some of those subscriptions, with library staff making some inconvenient changes in how they do their work in order to help reduce expenditures. That’s part of our own contribution to controlling expenditures.

After you’ve created the prioritized list and have the College’s final budget numbers for next year, can we weigh in before you make your final decisions on what to cancel?

We wish we had enough time. Unfortunately, we need all of your input well before that. We won’t have the numbers for our acquisitions budget until after Commencement, and the cancellations have to be communicated to our database providers in time for the changes to take effect July 1. They often want 60 days’ notice, which we won’t be able to provide in any case. We will definitely share the information about what has been cancelled and why.

Why didn’t you start sooner?

We were planning a different review for this year (single-title journal subscriptions) to give us some budgetary room to fulfill the requests we frequently get for new journals. We shifted gears and started on this much larger review as soon as we learned that we would be asked to plan for deeper cuts; that initially planned review would not have provided enough in cuts to meet the College’s goal.

What does "usage" mean?

Usage is the number of online views or downloads of full-text units (e.g., a single journal article, ebook chapter, full ebook download, streaming video file, image, etc.), for the most recent full year available. Most of the numbers provided are from July 2015 - June 2016. Not all database providers make that information available, and some of it has limitations making it misleading when viewed alongside other databases. Those data are not presented here; only the directly comparable numbers are.

I see that some databases have low use and high costs. Is there a cost-per-use limit?

No. Some academic libraries do use a firm cutoff point for cancellation decisions. We feel that approach is more rigid than is appropriate for Middlebury, so we consider cost per use as just one data point in a range of considerations. As outlined above, different disciplines use materials differently, and databases in the humanities are usually less expensive than those in the sciences. Some databases are critical resources for a given discipline. All told, cost per use is important, but it’s far from being the only criterion we want to consider.

How will you make the final cancellation decisions?

Library staff will analyze the input you provide on the importance and value of the various databases, and then consider it along with factors including the following (listed in no particular order):

  • Importance to a given discipline (i.e., the degree to which it’s a standard source).

  • Uniqueness of content (i.e., whether the information it provides is also available via another source to which we have access).

  • The degree to which we might own the material (i.e., whether we paid a one-time fee for content when we first subscribed, whether we own only some of the content, etc.).

  • The degree to which the material might be available through other sources (e.g., interlibrary loan).

  • Whether we have already cancelled other, similar databases in the past (especially those focused on the same discipline).

  • Cost per use.

  • Historical inflation, and our assessment of likely future inflation.

  • The number of people (students and faculty) and the type of work likely to be affected.

In developing the final prioritized list, we’ll aim for the best balance between savings and overall impact. Once we have the budget numbers, we’ll see how far down the list we’ll have to go in order to meet that budget.

Will you share the prioritized cancellation list before the cancellations are actually made?

We’ll try! We hope to have all the data analyzed before Commencement and before people leave for the summer, but that’s not a certainty. The timing is tight.

About the Survey

Who can give feedback?

Anyone with a current Middlebury email address who receives the survey link via that email account. It includes faculty members associated with Bread Loaf or Language School programs.

How long does the survey take?

If you review the databases in one discipline, it might take 10-15 minutes. (You will be invited to review more than one discipline.) If you don’t have enough time to finish, don’t worry! You can close the survey and return to it later. Just revisit the link at the top of this page. The subject breakdown:

Humanities: 42 databases

Social Sciences: 34 databases

Sciences: 26 databases

Multidisciplinary/general: 44 databases

Many of the general databases have deep material in multiple fields, so they’re not necessarily things you’ll want to skip over.

Is the survey anonymous?

It is confidential, but not anonymous. Only a very few library staff who compile and analyze the results will see any identifying information, which will be stripped out before those results are presented (including where the response itself might identify the respondent). We ask for it in the first place so we can follow up for more information on comments or questions, and to assess response rates in the various departments. The latter piece of information is also important in our evaluation; we want to know whether the electronic databases in that subject are important enough to devote the time to the review.

How do I know which package has a particular journal?

You can search our journal holdings from the library's main page; simply click on the "Journals" tab and enter the title. The search results will tell you what packages offer that title.

Can I review more than one discipline list?

Yes. At the end of the feedback instrument, you can choose to end the survey and submit it, or you can return to the beginning and select another list. You can also submit the survey and come back to it later on to give us more input.

Why do this through a complicated survey when we could just talk to our department liaison librarians?

You can certainly talk with your liaison, but there are a couple of reasons for doing it through the survey: First, we’d like to gather the feedback systematically; if we can get input in the same format from everyone willing to contribute, the results will more fairly represent everyone’s needs. Second, faculty meetings can be difficult venues for squeezing in a lot of information. It’s frequently difficult for librarians to get on department-meeting agendas so they can address the whole group with the same information and answer questions for everyone at once.

Can I take the survey more than once if I feel really strongly about it?

Yes. It’s lengthy, but if you feel strongly enough to devote the time, that’s useful information for us.

Who can I talk to about other questions?

Douglas Black, Head of Collections Management, is the person responsible for the survey and, ultimately, for the collection decisions. He’s happy to hear from or meet with anyone. You can also contact your department’s liaison librarian, who will be able to answer your questions directly or will help you connect with Douglas.

JavaScript disabled or chat unavailable.