Movies, TV Shows, Songs, and Textbooks

You can stream just about any kind of content on demand now. Rather than purchasing a single DVD or CD, companies like Netflix and Spotify have popularized a business model where customers pay a monthly fee and get on-demand access to a huge library of content.

Now, everyone knows that the college textbook market is horrifically broken. But just how “there’s no way that can possibly be true” broken is it? Here’s a quick comparison between major media types that shows how INSANELY EXPENSIVE textbooks are.

Movies and TV Shows

Amazon Prime – $6.59/month ($79/year) for access to 10,000 movies and TV shows
Netflix – $7.99/month for access to 20,000 movies and TV shows
Hulu Plus – $7.99/month for access to 45,000 movies and TV shows


Spotify – $9.99/month for access to 15 million songs
Rhapsody – $14.99/month for access to 14 million songs


CourseSmart (“world’s largest provider of digital course materials,” sells digital access to other publishers’ textbooks)
– $20.25/month ($121.49/180 days) for access to one biology textbook
– $18.25/month ($109.49/180 days) for access to one world history textbook
– $18.49/month ($110.99/180 days) for access to one algebra book

Online, on demand access to one textbook (~$19/month) costs more than online, on demand access to every major movie, TV show, and song produced in the US in recent memory ($7.99 + $9.99 = $17.98/month). Really. One textbook costs more than the entire output of the film, television, and music industries combined.

Go on… read that previous paragraph again. The math is correct, but it’s not right. Not in any moral sense of the word “right,” anyway.

Amazon Prime and Netflix data, Hulu Plus data, Spotify data, Rhapsody data, Biology book data, Algebra book data, World History data

P.S. I’d love a great visualization of these data. If you feel inspired to create one, post a link to it in the comments below.

6 thoughts on “Movies, TV Shows, Songs, and Textbooks”

  1. This is appalling. However, isn’t this highlighting a problem with the model we are using? Why are we still directing students to an “expert” in their learning instead of asking them to compile something themselves. I’m sure that a student assembled resource (textbook) done by a biology (or history or algebra) class under the guidance of a lecturer could be as well presented as one rented. And, the learning experience of the students putting the resource together would be so much better than the passive read and test model that passes for learning today. (Jesse Martin –

  2. I would agree that the textbook industry is broken – but there are some incomplete aspects of the analysis:
    1) There is still an economics of scarcity – but it’s a scarcity of attention that should drive the costs down. If anything, textbooks should compete for students’ time by being efficient and effective learning tools. Authors would get paid based on the fraction of usage their learning objects command from the total usage of the whole subscriber base across all the content in the subscription pool. This competition should drive down the price for learning objects by giving an incentive for publishers to make it up in volume from every student learning a subject (as opposed to the fraction whose professors have exclusively adopted a single textbook).

    2) On the other hand, the temporal element is missing – which should drive subscription prices higher than consumer content. For example, a movie has about two hours of experience. A novel has perhaps 4-8 (depending on reading speed). A textbook is a large dense piece of content that might have tens to hundreds of hours of “enjoyment”. Calculated on a per-minute basis the monthly costs aren’t too bad – assuming that you had an omnibus “per-subject” (or better yet an “all subjects”) subscription. Textbook subscription services could be larger if there were more alternatives competing for the student’s time within the subscription.

    3) Third, like drugs and movies, the true pricing is tiered. Early adopters pay a significant premium for early access or access within a favored distribution mechanism. A first run movie is $12+ (and up to $18 if in IMAX/3D); this high initial price covers part or all of the production cost so that later viewers only have to pay a fraction of the price (collected from netflix and passed on to the studio). Music from headline bands might be subsidized from a live concert tour. ebooks are subsidized by the initial hardcover release (and sometimes a paperback release). Tiered pricing works well when the total market is large and spread out over time (since great movies, books, and music can be sold across generations). I don’t know if it’s possible to provide a tiered model in education such that early customers cover plant costs and allow later customers to pay far less. I think we have to drive down the initial production cost be removing the redundancies that currently require every textbook to provide full coverage of all topics (i.e., break the assumption of a single textbook exclusively used by a schools).

    4) Finally, there needs to be some room for “value-based pricing” in an academic content subscription. Watching movies has entertainment value, but doesn’t increase a student’s earning power. Learning content that just suts there and lectures isn’t necessarily more valuable either. However, if learning objects offer the ability to earn recognized badges and other objective measures of competency (as you so ably write in your blog postings about jailbreaking the transcript), then acadmic subscriptions could command a higher price than entertainment subscriptions. Perhaps we could go further and charge a much higher price for earning the badges and credentials – no matter how many learning objects or how much time a student chooses to spend.

    The last point is an interesting way to align the goals of learners with those of providers. Learners want “competency-generating density” and providers want “revenue-generating density”. The more effective the learning objects and the supporting environment in which they’re consumed – and the more effective the students become at generating evidence that they have learned and can create – the more value is truly created and exchanged.

    These are just some off-the-cuff thoughts. The textbook industry has to get much better at building smaller units of learning in shorter amounts of time without the artificial cost burdens of freebies for instructors and door-to-door sales reps trying to secire exclusive adoptions. This probably means there needs to be a wholesale replacement and disruption of the current providers (leopards being generally unable to change their spots).

    Thanks for calling to our attention the peculiar economics of the academic content industry (currently delivered as textbooks) vis-a-vis the economics of other content industries.

  3. So I am no apologist for the textbook industry or their exorbitant prices, but if we are going to be fair, it seems like there should be at least two additional columns, average cost of production and total potential audience. This doesn’t undercut your argument, but it feels like it would be more honest if we at least compared apple-like things.

Comments are closed.