The Indiana University eText Experience: The Economics
Brad Wheeler, IU Vice President for Information Technology and Chief Information Officer; Professor of Information Systems, IU Kelley School of Business
Count me as optimistic that we can dramatically reduce the costs of required, digital course materials in 2018 and beyond. The publishers are now aligning to the compelling win-win economics of eTexts and “Day 1 Access” models at an accelerating pace.
As we all celebrate this favorable development, we must stay focused on addressing the root causes of the dysfunction that got everyone into the sorry state of high priced books. The history of how we got to this moment is instructive as students, faculty, staff, authors, and publishers work to re-envision and implement a rational model. Before 2018, the main family feud for pricing was within the publishing companies themselves, but now the feud is more acute within our colleges and universities. The reasons for the dueling feuds within the publishers and within our institutions are remarkably symmetrical, and each can be resolved by letting go of legacy sales and distribution models that have long outlived their economic value.
This chapter addresses the role of contracts, negotiations, incentives, and the purposeful reengineering of an industry in the age that increasingly favors digital course materials. It chronicles some of Indiana University’s (IU) lessons from 2009 forward and the intransigent, structural reasons that have taken so long to get to a new model. It concludes with guidance for institutions that are ready to reduce the cost of attendance. It addresses contracting for paid course materials, and I’ll affirm here again that Open Educational Resources (OER) can be a powerful means to reduce the cost of attendance when available. Yet, until these resources are able to meet more course needs, we must concurrently work to address the prices of course materials from publishers.
Buyer and Seller without Middlemen — 2009-2011
IU envisioned a model where students paid less and creators got fairly paid for each use of their content. By creators, we meant both authors and publishers, and we sought to reduce the number of middlemen seeking to get paid by adding additional profit to the price of digital course materials. We did the modeling and had a strong case to present a new model that essentially “moved the toll booth” to the university bursar so that creators would get paid for each enrolled student who used their digitally delivered course materials.
In the spring of 2009, IU requested and received visits by delegations from the big three publishers (Cengage, McGraw-Hill, Pearson), who accounted for more than 64% of the dollar volume of all required course materials at IU (the big five commanded 72% of the dollar volume). All were very quick to explain to us that they were no longer “publishers,” rather, they had become eLearning companies. We said great, we want to buy a bunch of that in digital form—whether digital versions of paper books or adaptive exercises. Give us a great deal, and we’ll pay you for every use of your content. Sounds great…right?
We viewed the deal as having two parts. First, we only wanted one eReader/Annotation piece of software to promote and support for students and faculty. We thought it a bad idea if a freshman had four different pieces of software just to read five different books. Thus, we licensed the software at IU for use with any content from a third party. Second, we rigorously negotiated with the publishers to give us the lowest possible price with no embedded software or distributor costs. From late 2009 through early 2011 IU subsidized four semesters of small but growing trials with some content also contributed by publishers as we all worked out the plan.
Then came the formal Request for Proposal in March 2011. We were all ready to march forward with big new deals as we had support of students and faculty, positive pilot trials, bursar billing, and all was ready. We sat by the mailbox and waited for the responses. And waited. And waited.
What? How could the publishers not accept a collective increase in revenue for their products over the current failing traditional model? We did receive some dead-on-arrival responses. They sought price points in the 50-75% of list price range and heralded a “50%” discount off the made up and often ridiculous list prices of books. As noted in earlier sections of this book, the run up in list prices was largely driven by seeking greater revenue from an increasingly failing model that only addresses part of the market. IU sought flat pricing or a subscription deal for an entire publisher’s catalogue, but that was a bridge too far (until 2018). The entire industry was premised on maximizing list price and then discounting to wholesale bookstore distribution channels.
We were at a stand-off that lasted for several months. Finally, Tom Malek, then with McGraw-Hill and now with Pearson, met me in Indianapolis and we cut a deal at 35% of list price (65% discount). Four smaller publishers were ready, but they would not move without one of the majors going first. Pearson walked away but came back and took the same terms the next summer. Cengage took another two years as they kept holding out for a higher price to students, and we just left them to the retail market without getting paid for each use of their content. A turnover in executives there brought them in quickly to sign on to the same terms. Our adoption numbers accelerated, and we eventually signed 25 publishers. Elsevier, law, and other health publishers remain holdouts in early 2018, as the structure of the market for their books to support post-graduation credentialing exams (the Bar Exam, Nursing License, etc.) tends to have the characteristics of a more biased monopolistic or oligopolistic market. Thus, unlike other publishers, they face less competition and have fewer real incentives to improve their pricing to students who have very few choices for substitutes.
Throughout this period and to today, IU does not mark up the costs of the materials that it negotiates for on behalf of students.
Economics and Incentives — 2012-2016
One of our early lessons is that we were not negotiating with an individual at a company. Even when the individual was a very high-level executive, we slowly came to understand that we were really negotiating with an internal family feud. The incentive structure and bonus compensation for various sales executives, product managers, division heads, etc. was a complete mishmash of unaligned incentives in a firm. Consider the following example of the traditional textbook sales model:
If there are 100,000 students in the US taking an Introduction to Marketing course, and the most popular book has a 75% market share, you would assume expected sales of around 75,000 books a year. That book lists for $187 and the publisher charges various book resellers around 60% of list or $112. In the first year of the new edition of the book there are no direct substitutes, so if 95% of the students bought the book, publisher revenue would be 75,000 x 95% * $112 = $7,980,000. In year two, there would be an abundance of used books, so new sales may fall to 30% of the market ($2,520,000), and by year three down to 15% of the market ($1,260,000). The three-year maximum revenue to the publisher would sum to $11,760,000, but publishers have to reduce that number by returns from resellers, sales commissions, author royalties, printing and distribution, and other cost of sales deductions, so let’s generously call it $10M in net revenue on 105,000 actual book sales in a potential three-year market of 225,000 students.
Since the costs to write, edit, and publish the first copy of a book in digital or print form are largely the same, the best economic analysis is on revenues and costs for all sales after the first book. If a publisher’s cost to print, warehouse, and ship a book is $10 (and it is much lower than that), then a wholesale price of $112 appears to have a gross margin of $102 per book. Thus, every year, the publisher’s financial statements would show it has a continuing strong gross profit margin on each sale, though total revenue is falling quickly as students rationally seek substitutes in used books or other means. Used books are a perfect substitute for buying a new book, and since they are much less expensive, students are strongly incented to seek substitutes for new books in the traditional model.
Now consider a different approach. At what price would a publisher need to sell the book in an eText model where 225,000 students each were bursar-billed for the book? $10,000,000 / 225,000 = $44 per student to achieve the same net revenue to the publisher. Win-win. The publisher got the same revenue, all the students have the course materials, and the cost fell to 23.5% of the made-up list price.
The financial statements for the publisher, however, would look much worse compared to prior years as the gross profit margin falls to $44 less cost of digital distribution (say $5) yielding $39. This is far short of the appearance of $102 gross margin—a reduction of 56%. Of course, this is all quite ridiculous: By driving for a high margin on a diminishing part of the market each year or a smaller margin on the whole market, the publisher is in the same net cash position. Yet that was just part of the internal family feud at most publishers. Bonuses were tied to the traditional model metrics, and that drove behavior.
While either of these scenarios yields similar profit outcomes for the publishers, the total costs to students could not be more different. In the traditional publishing and distribution model, students switch to used books and resell over the three-year cycle. Our analysis shows that the 225,000 students in the traditional model, including used books and sell-backs, would have paid over $21,000,000 or an average of $95 per book—though publishers only receive a fraction of those funds because of the used book market, reseller fees, shipping, and other costs of goods sold. Students in the eText model would have paid a total of just over $10,000,000 , or $45 per book, and nearly all funds are transferred to the publisher, with substantial savings to the student. In both cases, the publishers received the same total money, though students spent less.
IU spent much of 2012-2017 working with the revolving set of publishing executives as we kept trying to enroll more publishers in this win-win model for both content creators and content consumers. We had publisher sales representatives calling on faculty at IU and urging them not to use the eText contract that IU had signed with their company. It was crazy, yet our adoption numbers and (pass-through) revenue in the eText program continued to accelerate.
This period also saw a rise in textbook “rental programs” and so-called eText rental programs, which allow a student to rent a physical book or eText for a set period of time, normally one semester. I view both of these as marketing bandages on a wounded economic model rather than longer-term solutions. Rentals are no more than a marketing label for new/used books with a guaranteed buy-back at a known cost of ownership. They absolutely do not address the root cause of the failing traditional model, do not pay content producers for their work, and do not achieve the lowest possible prices for students. Likewise, the concept of an “eText Rental” has nothing to do with the real costs and educational value in an industry. Why should we encourage students to take a lower-cost, one semester eText Rental if the course spans a whole year? The real cost of providing continuing access to digital content for a few more days or years is negligible. I view both of these as highly distracting Siren Songs that reward 3rd party resellers over students and institutional goals.
This period also saw a maturing and growing popularity of Adaptive software for courses (e.g., Pearson MyLabs, McGraw-Hill Education ALEC, etc.). Adaptive content is cloud-hosted software by a publisher, and its development and operational costs per student are higher than a physical book or digital version of the same book. Many publishers were making extraordinary investments in building the so-called “Robo-tutors” as a strategic bet. With only a single provider, these systems took on the pricing characteristics of a monopolistic product — students had no real substitutes if these were required for a course and homework was graded by them.
Throughout this period, IU renewed its contracts with about 25 publishers, and in each case, we were able to demonstrate reliable, growing, and aggregate sales volume to gain slightly improved pricing each round. We included Adaptive in our renewed contracts, and often that was at a flat price with some discount relative to any other means of students gaining access to it.
The World Changed — 2017
Then the world changed. Shares of Pearson PLC had already lost 50% of their value from early 2015 to late 2016. By early January 2017, they had shed another 20% of their December 2016 high, and this period portended a rapidly accelerating shift in the mindset of most of the publishers. All of a sudden, publishers were now aggressively driving—not acquiescing to—the eText model and began branding it as “All Students Acquire” or “Day One Access.” Both of those terms appeared at multiple publisher booths at the 2017 EDUCAUSE Annual Meeting in Philadelphia. More executive exits, shifting political victories within the publisher family feud, and greater willingness of institutions to act on bursar billing were all fueling the change. The major campus bookstore chains had also rolled out versions of the program, but they tacked additional profit margin on each book for their middleman status that again upped the costs to students.
These changes in the publishers’ mindset coincided with maturing acceptance of digital by faculty and students and growing pressures to reduce the cost of attendance. Major online book sellers and rentals were also putting pressure on the industry for change. Several of the publishers also moved to an “Agency Model” where they stopped selling books for selected courses. Instead, they would provide those books to distributors who each rented them out to students at a maximum set price for which the distributors received a publisher-paid commission on each rental. Demand for Adaptive course materials accelerated, particularly for large general education courses.
Peace within both Families 2018-
It has never been easier than now for an institution to assertively reduce the cost of attendance by helping its students acquire digital course materials at the most favored prices. Unfortunately, what I often see is an abundance of administrative action that fails to actually seek the most favored prices for students as colleges and universities acquiesce to other distracting objectives that yield higher-than-necessary costs on students. Thus, as new leadership at many publishers has made peace with their internal family feud, it is also time to do so within the academy by affirming our twin goals to reduce the cost of attendance and improve learning outcomes (covered in following chapter The Benefits of eTexts for Students and Instructors) and then pursuing them without distraction.
I see four paths forward in order of decreasing attractiveness:
- Participate in a consortium to manage the contracts and aggregate scale
- Negotiate directly with each publisher (as IU had to do in the early days)
- Access the model through resellers/bookstores at a higher cost to the students
- Do nothing and let the retail model prevail
The first three paths necessitate that an institution be willing to collect course material costs directly from students via bursar billing. It is the essential trade (almost) every student acquires and pays in exchange for substantially lower course material costs. Most institutions have long had various course fees, and the technical lift to establish these fees for eText and related policies is now well understood. Choosing a path for contracts is the next hurdle.
If a marketplace has 10 providers and 10 consumers, a set of individual relationships would yield 100 separate contracts, and that would not be efficient for anyone. Thus, to continue to go it alone in option #2—as IU did in our founding of the eText program—is not advised. Even the largest amongst us has trivial scale in the whole of higher education. IU chose this path as none other was available at the time, and we were willing to invest substantial negotiating and contracting efforts to establish the model. I am grateful to Jill Schunk, Associate Vice President of IU Procurement Services, for her help and tenacity as we pressed forward with direct negotiations. IU is quite fortunate that wise people before me chose to exclude digital content from our outsourced bookstore agreement in 2007.
By 2017, IU was assertively moving our eText agreements to the member-owned Unizin Consortium (Unizin.org). Unizin has the scale of over 800,000 students and makes an efficient basis for publishers to contract with Unizin as an entity and for member institutions to contract with Unizin. As a not-for-profit entity, Unizin is an excellent aggregator for its members, and there is no reason that other state consortia or other alliances couldn’t also aggregate scale for an efficient execution of option #1 above.
The outsourced bookstores path (#3) also aggregates scale for contracting with publishers. Since they are a for-profit business, they rightly choose to make a profit on course materials in all models and pass that cost on in higher prices to students. At some institutions, they may also serve as a transfer mechanism for money from student purchases to various accounts within the university. If profits on student course materials are a key institutional objective in the form of a transfer payment from books to some other worthy program, then this could also be transparently achieved through other mechanisms of a university mark-up on eText prices. The challenge, of course, is that adding cost on course materials to transfer payment for other uses works against the goal of reducing the cost of attendance.
The most worrying path is the default inaction of #4. By definition, it means that inaction by institutions subjects students to the retail pricing market. Institutions lose all real possibility of aggregating learner analytics across publishers, and students will be compelled to pay more than is otherwise achievable in the win-win eText model. Staying with the traditional model also incentivizes students to consider not buying the course materials or to use prior editions. These choices generally work against improving learning outcomes, and they are quite avoidable with institutional action.
Finally, there are the terms of the contract itself whether negotiated by a consortium, the institution, or a reseller. At minimum, and in our experience, a deal with a publisher should include all of the following on the most favorable terms possible:
- Highly favorable prices relative to any other means of legally accessing the material.
- Ongoing access to the content for at least as long as the student is enrolled at the institution.
- An ability to print.
- Access through an institution-selected common eReader/Annotation software to reduce support and training costs at an institution.
- All data from clicks of reading or working through Adaptive Tutors available to an institution.
- Ease of management for allocating student access within courses and sections to digital materials.
- Efficient and quick access to unlocked PDF copies of textbooks, as needed, for university assistive technology and accessibility centers to accommodate textbooks for students with disabilities.
We have achieved these in IU’s early years, and we are now achieving them through our consortium. Your institution can too. And it should. It is win-win for both content creators and content consumers.
 For more on “Robo-tutors” see: Steve Kolowich, “Score One for the Robo-Tutors,” Inside Higher Ed May 22, 2012, https://www.insidehighered.com/news/2012/05/22/report-robots-stack-human-professors-teaching-intro-stats or Issie Lapowsky, “This Robot Tutor Will Make Personalizing Education Easy, “Wired August 26, 2015, https://www.wired.com/2015/08/knewton-robot-tutor/