on why introductory textbooks are so expensive

David Diez works on the OpenIntro team, which develops open-source textbooks, statistical software labs, and other course resources for intro-level statistics. This video above is the talk he gave at useR! 2014 entitled “Textbooks fail where Software Succeeds” in which he outlines the challenges unique to open-source textbooks, the state of OpenIntro, and how people can get involved. Below, he discusses these challenges further and offers his thoughts around their solutions.

Textbook prices have risen faster than tuition and fees at colleges, and they’ve even risen faster than health care costs. This is a clear signal to us here at OpenIntro that something is fundamentally wrong with the way the textbook industry currently operates.

OpenIntro and every open-source textbook developer face an entrenched and profitable publishing industry. We want to compete head-to-head with the traditional publishers in a free market, which we believe will generate better educational products at better prices for students. However, the intro-level textbook system has developed in a way that we think prevents this free market from existing for three reasons:

  1. Students, who are the market’s price-sensitive consumers, do not freely choose which textbook to purchase. That’s a job of each course instructor or department.
  1. The teachers and departments who decide which book students should buy aren’t impacted by textbook price: they typically receive free textbook copies called “desk copies” for teachers and teaching assistants from the publisher.
  1. Virtually no one other than college students buy new introductory-level textbooks, so there are few price-sensitive consumers with decision powers which influence the market.

These three realities form the foundation of why intro-level textbooks are so expensive today. If textbook prices are going to ever change in a free-market fashion, we believe that we will need to fundamentally change at least one of these three pillars.

To displace the first pillar, it would require teachers to give up the ability to assign a particular textbook for a course to students and instead rely on students selecting suitable course materials, perhaps by reading reviews and comparing prices. Instructors could also rely on open-source exercise repositories to avoid reliance on a single textbook for homework problems. However, these possible solutions are not without their challenges. For example, it would be difficult to ensure each student purchases a textbook that covers all of the required topics, and some textbooks also use slightly different terms and notation. Such problems are far from insurmountable, but the known solutions add some burden onto instructors’ workloads, including adjunct faculty who frequently teach many courses simultaneously.

The second pillar also has no simple, universal solution. The primary goal is ensuring that departments and instructors consider price as carefully as students consider price. There are a remarkable number of instructors who pay close attention to the downsides of textbook costs, e.g. low-income students frequently skip buying course textbooks, and these instructors will seriously examine low-cost or open-source textbook options. However, this is not a universal solution — at least not yet. Here are a couple ways to try to nudge more instructors and departments to move towards a price-sensitive mindset:

  • One option is to increase price awareness among teachers and departments by highlighting the damage done to learning outcomes when students avoid textbooks due to their high price. This is the strategy OpenIntro has generally followed, e.g. through blog posts and talks at conferences.
  • A second option is regulatory, either on the state or national level. For example, regulations could restrict publishers from gifting “desk copies” of textbooks and other materials to state college departments. (I’ll talk about how OpenIntro currently handles gifting of textbooks later on.) Stopping this textbook gifting would ensure, for example, that whenever a publisher comes out with a new edition, departments would be hit with an unexpected cost, just like the students currently experience. If this happened in a course that had one instructor and 4 teaching assistants, the cost of buying 5 copies of the new edition could reach $1000 or more. This could further motivate real consideration of lower-cost options.

The third pillar considers the possibility that there may be a third option: create a large, new market outside of the university system for introductory-level textbooks that is price-sensitive and that will be tempting enough for publishers to pursue. This new market would need to be a comparable size to that of college students to be effective. However, this solution has two problems: there is no such market (e.g. instruction of industry employees is a much smaller market), and many consumers in existing markets are already well-served by free online resources. That is, to break this pillar would require publishers to see the value in pursuing a smaller market where they will most likely have trouble competing, even when doing so will likely harm their current profit margin. (For some perspective, consider that a substantial fraction of the market for advanced statistics textbooks is from practitioners and researchers. That is, such books typically compete in a price-sensitive market, which often keeps prices of advanced statistics textbooks below the prices of introductory-level statistics textbooks, even though the market for each advanced book is much smaller.)

How is OpenIntro currently contributing to a solution for students?

We estimate that over 5,000 students (probably closer to 8,000 students) used OpenIntro’s open-source textbooks in 2014 courses, which amounts to $500,000 to $800,000 in student savings. For perspective: this is still less than 1% of the intro-level statistics textbook market, so we have a long ways to go.

Our open-source textbooks are downloaded about 10,000 times each month, and over 25,000 times each month during peak usage. OpenIntro is also sustainable, as our staff is 100% volunteer (and regularly expanding), and our costs only amount to about $2500 annually.

How OpenIntro textbook gifting works, and how you can contribute

Our textbooks are free online and physical copies are priced under $10 on Amazon (meaning they are sold royalty-free.) In 2013, we decided to start giving instructors up to 2 free OpenIntro textbooks. Since we have suggested possible regulation around such textbook gifts, that sounds hypocritical, right? Well, there are two big differences between what we do at OpenIntro and what traditional publishers do.

  1. Our textbook prices are typically 5-10% of the textbook price of our competitors. That is not a typo. Our prices differ by an order of magnitude, and we also give electronic versions away to everyone for free.
  1. Much more importantly, the “free” copies that publishers provide to teachers are indirectly paid for from the profits of textbook sales to students, which we think should change. In OpenIntro’s model, we have no profits — from students or anyone else — so we pay out-of-pocket for teacher copies since we know getting our books in the hands of teachers will ultimately help students save money.

OpenIntro’s model ensures there is no financial conflict of interest, so we think that sending each teacher a $10 textbook or two is quite reasonable.

We also now welcome Amazon gift cards at to offset our out-of-pocket expenses. These funds will be used exclusively to buy textbooks for teachers, or if we get an overwhelming response, to also buy textbooks for high schools and low-income students. If you’d like to help, consider sending an Amazon gift of $10. (Please note: gift cards are not tax-deductible.)


I appreciate the opportunity to write a blogpost for DataScience.LA, especially on such a controversial topic. A big thank you to Mine Çetinkaya-Rundel and Andrew Bray of OpenIntro, who provided great advice and substantial feedback on this post.

Editor’s note: The opinions expressed in this guest post do not necessarily reflect the opinions of the DataScience.LA Editorial Board, and we thank David for contributing this post and helping to stimulate community discussion around this topic.

Share This Post

Leave a reply