Market and Economics

Introduction

I think it's fair to claim that current copyright laws heavily influence the price of electronic resources, and in this section, we cover the basics of copyright, how copyright creates monopolies, and how those monopolies, in the digital era, are able to demand substantial sums of money for electronic resources, and finally, how this impacts library budgets.

Below I'll discuss copyright and the first sale doctrine, I show how digital works have disrupted some basic ways that libraries function. Then I'll discuss the impact that this law has on complicated e-resource collections and costs.

Copyright

Copyright law grants a monopoly to the person or corporate owner of an intellectual property. That is, copyright owners have exclusive rights over the material that they own, where the owners may be a person or an organizational entity. Section 106 of the law grants copyright owners the following rights:

(1) to reproduce the copyrighted work in copies or phonorecords;

(2) to prepare derivative works based upon the copyrighted work;

(3) to distribute copies or phonorecords of the copyrighted work to the public by sale or other transfer of ownership, or by rental, lease, or lending;

(4) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and motion pictures and other audiovisual works, to perform the copyrighted work publicly;

(5) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and pictorial, graphic, or sculptural works, including the individual images of a motion picture or other audiovisual work, to display the copyrighted work publicly; and

(6) in the case of sound recordings, to perform the copyrighted work publicly by means of a digital audio transmission.

Source: Copyright Section 106

These exclusive and all-encompassing rights are designed to allow copyright owners a monopoly of their property. While it is good to motivate the creation of intellectual property, without restrictions, there would be little to no benefit to the public. For example, if those exclusive rights were followed without limitation, then it would mean that the exchange of money for a work between a copyright holder and a buyer for something like a printed book or a DVD would not entail a transfer of ownership of that physical copy; that is, it would not allow the buyer of the physical item any distribution rights of the item once the first exchange has been made. Under such a scenario, libraries would be able to buy physical books but would not be able to lend them.

To address this, the First Sale Doctrine (also see the Justice Department's explanation) helps avoid the issue granted by the list of exclusive rights listed in Section 106. Because of the first sale doctrine, made precedent in the early 20th century and then codified into law in 1976, you, I, or a library may buy a physical copy of a work, like a book, a DVD, a painting, and literally own that specific copy. First sale doctrine does not grant reproduction rights, as listed in Section 106 of the 1976 copyright law, but it does allow anyone to distribute the singular, physical representation or embodiment of the work that they have purchased. Thus, the first sale doctrine is why libraries were able to thrive throughout the 20th century, lend material, and preserve it. More mundanely, it's also why I can buy a book at a bookstore and later give it away or sell it to someone.

The digital medium makes things messier, as it tends to do. There are two big reasons for this. First, digital works are not subject to the same distribution constraints as physical works are, and the first sale doctrine is about distribution rights and not reproduction rights. If I have a physical copy of some book and give you my copy of that book, then I no longer have that copy. However, if I have a copy of a digital file, then as we all know, it is relatively trivial to share that file with you without losing access to my own copy. Since digital works can be copied and distributed without losing access to the copies or to the original, the First Sale Doctrine does not necessarily apply to digital copies. Consequently, in the digital space, there are fewer limitations on supply, including on lending.

Second, many digital works are like software, or at least, they are intertwined with the software and hardware needed to display them. This is true for all kinds of documents, like HTML pages, which need a web browser or a text editor to read them; or audio files, which need a media player to listen to them. But consider ebooks as an example. Whether a printed book is the size of a quarto or as large as a folio, whether it is all text, whether it includes images, or whether it has pop-ups makes no difference to its basic distribution potential and ergo its copyright status. These physical works are, in essence, self-contained. Ebooks, however, arrive in all shapes and sizes. Project Gutenberg distributes public domain ebooks in various file formats that include plain text documents, that have no presentation markup like bold, italics, and like, HTML documents with markup, XML documents like EPUB, and then also PDFs and others. Recently, they started using AI to convert their books to audiobooks Why so many file formats? Text is text, right? In the print space, a book is simply text printed on pages of paper, even if it is sometimes printed on different sized pages or using different type settings. But these various markups exist because they each offer technological or presentational advantages and are often tied to specific pieces of software and hardware.

This is especially true for proprietary, encrypted file formats, like the ones that Amazon created for use only on Kindles (other e-reader and other stores also), or the popular MP3 file format for audio recordings that only recently became patent free. While file formats like these may not be necessarily counted as software, depending on how we define software, it is certainly true that file formats and the specific software applications that display or play them are intertwined. If you are old enough, you may remember the headaches caused with files created as .doc in some early version of Microsoft Word that later failed to display properly in a future Microsoft Word version or in some other word document software or on some other operating system. WordPerfect 5.1 was a popular word processing applications in the 1990s, and it's not clear if files created with that application, or other popular word processing applications at that time, would open today, at least without intervention. In short, these complexities introduce obstacles to the first sale doctrine and raise other copyright issues because of the connection to software, which is also often copyrighted.

The main idea here, though, is that copyright holders and publishers have little financial interest in selling actual digital copies of works since they cannot prevent future distribution without special technologies. Instead, they are motivated to license material, and retain ownership, and sometimes explicitly tie that material to specific pieces of software and hardware, like the Kindle, which would have to be bought. That, we should note, adds additional expense.

Take note of the recent lawsuit against the Internet Archive (IA) in Hachette v. Internet Archive. In the early days of the pandemic, the Internet Archive started lending books through OpenLibrayr.org that were mostly not available as ebooks. The IA only lent these books if a library owned a physical copy and limited lending to the amount of physical copies owned. This is called Controlled Digital Lending (CDL), which the IA argued was Fair Use. Four publishers sued the IA and won the initial suit in March. The decision will be appealed but has already emboldened other media companies to sue the IA. But at stake is the basic notion of whether ebooks can be bought or must be perpetually licensed, until at least they enter the public domain.

Impact on Ebook Collections

What does this mean for libraries in the digital age? It means that libraries buy less and rent or license more, and renting means that they continually pay for something for as long as they want access to it. As Sanchez (2015) puts it,

At its simplest, this takes the form of paying x dollars per year per title during the length of the contract (Forecasting sect, para 4).

When the total supply of works increases, e.g., the total number of published books increase, as they do each year, then it means renting more and more without ever completely acquiring. (It also means, holding collections at a stable number, providing access to less.) When budgets are cut or remain stagnant, this ultimately entails a decline in the collection a library has to offer, or if not a decline in the collection, then cuts in some other areas of a library, like the number of librarians or other staff. This is the conundrum that Sanchez raises in his article.

If that alone were the issue, maybe librarians could discern other sustainable ways to proceed, but Sanchez (2015) raises additional issues and questions: what if publishers raise the prices for digital content at an annual rate faster than what they already raise for print content (reasonable assumption)? If so, does that mean that librarians will be able to afford fewer titles, digital or print, unless they raise their budgets, and, as they weed, how would that impact the physical space of the library? (See figure 2.3, specifically, from Sanchez's article. The plot shows just how much could be lost and how little gained if the forecasts Sanchez discusses come true.)

There are many ways to put constraints on the supply of an item in the digital landscape, as opposed to limiting supply in the physical space, which involve fewer methods. That is, it's relatively easy for publishers and others to restrict the supply of physical works. They simply have to limit how many of those physical works are manufactured (e.g., the number of print runs). But given the nature of digital content, restricting supply is driven by the technologies available to do so, and since there are so many publishers and distribution points, then each one of these points will often create their own unique type of constraint on the supply. The result is that there will be a number of confusing methods implemented to limit constraint, even if these limitations are marketed as selling points. In practice, this may mean that only a limited number of people may "check" out a work from a library at one time, or access a database at one time, and so forth. Thus, the budget issue has an impact on access and usability.

There have been recent attempts to address these issues. Paganelli (2022) describes some state by state efforts to lessen the financial burdens on libraries that e-content entails. However, as Paganelli notes, these efforts have largely not succeeded. And Brewster Kahle, the founder of the Internet Archive, has recently highlighted the growing attacks on library budgets that make this the outcomes for library budgets more pessimistic. From a publisher's perspective, Sisto (2022) argues that the general narrative about the tension between libraries and publishers is misleading. Instead, the author argues, the landscape is much more complex, and the publishers have made a number of attempts to "make their e-lending policies better for librarians" (2018-2019: Policy Updates and Different Opinions sect, para 6). Personally, I'm not sure I buy (or lease) many of Sisto's arguments, but I think one thing is clear: the e-lending market is complex and miscommunication abounds.

Read more about copyright:

https://www.copyright.gov/title17/92chap1.html

Impact on eJournal Collections

Although ebooks likely represent the biggest impact on public library budgets, academic libraries are largely concerned with scholarly journals. Like Sanchez (2015), Bosch, Albee, & Henderson (2018) show that the major issue is that academic library budgets are declining or holding flat even though prices continue to increase for journal titles and even though the number of published articles increase. This raises an interesting phenomenon: although researchers are hurt by the lack of access to research, researchers are also part of the cause of the supply.

The authors also note that part of the drive to publish includes a drive to publish in so-called prestigious journal titles, where prestigious is determined by how well cited the title is. The authors refer to a few citation-based metrics that the research community uses to determine prestige. These include the long-established Impact Factor, which can be examined in the Journal Citation Reports (JCR) provided by Clarivate Analytics, as well as newer ones, such as the Eigenfactor and the Article Influence Score, which can also be examined in JCR (the eigenfactor.org site is not well updated, at the time of this writing).

One motivation for using a citation metric as the basis of evaluating journal titles is because citation metrics indicate, at some level, the use of the work. That is, a citation to an article in a journal title means, hopefully, that the authors citing that article have read the article and used the knowledge from that work to add new knowledge. Historically, when Eugene Garfield invented the Impact Factor, it was partly as a tool for librarians to use in collection management because he recognized this use-based theory of citations.

However, citation metrics should never be the sole or even primary tool to evaluate research. While they may provide good information, there are many caveats. First, there are different fields of research, and some fields cite at different rates and at different volumes than other fields, and also for different reasons. This is why, in Table 5 of the Bosch, Albee, and Henderson (2018) article, the cost per cite for journals in the Philosophy & Religion category are so much higher that the cost per cite of titles in other categories. Authors in P&R simply have different citation and publishing behaviors than authors in other categories. Second, citations do not capture all uses of a journal. For example, there are many journal titles that I might use in my courses but may not use in my research, and this is true for other faculty, yet citation metrics won't reflect that kind of use. The authors refer to altmetrics, which was invented to help capture additional non-citing uses of scholarly products, but altmetrics is still in its infancy and is largely dependent on data sources and scholarly behavior that are problematic themselves. Third, there are various issues with the metrics themselves. The Impact Factor is based on an outdated calculation and is thus not a very appropriate statistical measure. The other metrics were created to address that but may have other problems. And four, the use of the metrics, regardless of which one, tends to drive publishing behavior---such that journal titles with higher metrics tend to attract more submissions and more attention, thus driving more citations to them. Such skewing drives demand to publish in those journals. Thus, citation based metrics are comparable to a kind of capitalist economic system where, as the sociologist of science Robert Merton noted, the richer get richer (in citations) and the poor get poorer. The issue then is that prestige, defined in this way, does not necessarily indicate quality: just use.

The authors also discuss some issues with Gold Open Access (OA) and the idea that Gold OA may compound the cost problem. This is where authors pay a publication fee, or an article processing charge (APC), once a manuscript has been accepted by a journal (there are other types of Gold OA cost models). We can do a quick off the cuff and rough calculation to see why this might compound the problem. As an example, PLOS ONE is one of the largest gold OA journals and in 2023, they charge an APC of $1,931. This is $336 more than what they charged in 2020, when I wrote the first draft of this section, which represents an overall 21.07% increase (see Table 1 below). In 2018, 32 papers were published in PLOS ONE that included at least one author from the University of Kentucky, totaling $51,040 in APCs for the 50 total institutions that were associated with these papers. This amounts to about $1020 per institution, in 2018 dollars, paid for by the authors and not libraries. For UKY authors, this also amounts to over $32,640 spent on APCs (32 * $1020). This is about $27K more than the average price of the most expensive category, Chemistry, as reported in Table 1 of the reading. So even if open access reduced costs to libraries, it still may not reduce cost on taxpayers, who fund much of of this research.

The 2022 public access mandate issued by The White House Office of Science and Technology Policy might make things more interesting. The OSTP memorandum states that "federally funded research must be publicly accessible without an embargo on their free and public release." This mandate requires that federal agencies with research and development expenditures to develop their own policies, and that could mean that such policies result in Gold OA agency-specific mandates or Green OA mandates. Green OA allows pre-prints to be made available (article versions before peer-review) or post-prints (article versions after peer-review) but not publisher versions (versions after formatting, etc.). We'll see how this plays out.

Readings / References

Bosch, S., Albee, B., & Henderson, K. (2018). Death by 1,000 cuts. Library Journal, 143(7), 28–33. https://www.libraryjournal.com/story/death-1000-cuts-periodicals-price-survey-2018

Sanchez, J. (2015). Chapter 2. Forecasting public library e-content costs. Library Technology Reports, 51(8), 9–15. Retrieved from https://journals.ala.org/index.php/ltr/article/view/5833

Additional References

Paganelli, A. (2022). Legally Speaking—States Unsuccessful in Providing Financial Relief of eBook Terms for Libraries. Against the Grain, 34(3). https://www.charleston-hub.com/2022/07/legally-speaking-states-unsuccessful-in-providing-financial-relief-of-ebook-terms-for-libraries/

Sisto, M. C. (2022). Publishing and Library E-Lending: An Analysis of the Decade Before Covid-19. Publishing Research Quarterly, 38(2), 405–422. https://doi.org/10.1007/s12109-022-09880-7