Recreational books have deflated in real terms since 1997 (-0.09%/yr CPI), while publishing EBITDA (~13%) sits below the 16.56% market average, near grocery-store margins.
Key Takeaways
BLS CPI data: $20 of recreational books in 1997 costs $19.49 today; books are the only flat line against housing, healthcare, and entertainment costs.
Inflation-adjusted, To Kill a Mockingbird and Fellowship of the Ring should retail for $43-54; actual 2023 hardcover prices are ~$27-28.
Publishing EBITDA (~13.18%) trails semiconductors (36.77%), software (35.93%), pharma (33.59%), and even alcoholic beverages (29.52%).
Retailers take roughly 50% of cover price; a $2 cover increase nets the publisher ~$1/unit, meaningful only at 15,000-30,000+ copy volumes.
The returns system lets bookstores send unsold inventory back for full credit, leaving publishers with zero revenue on returned units despite bearing all production costs.
Hacker News Comment Review
Commenters pushed back that the monetary framing ignores physical carrying costs: shelving, larger apartments, and moving loads compound over years of book acquisition.
Strong consensus that college textbooks (Pearson, Wiley, Cengage, McGraw) represent a separate publisher-monopoly problem the article deliberately sidesteps.
The “books are too long” counterpoint drew clear support: publisher and retailer incentives encourage padding 50-60 pages of real content to justify $20+ price points, diluting per-dollar value without reducing price.
Notable Comments
@WillAdams: A tax law change required publishers to pay taxes on unsold warehouse inventory annually, effectively killing the backlist model and forcing smaller initial print runs.
@hgoel: Consumer willingness-to-pay is about certainty of enjoyment at purchase time, not CPI math; ebooks solve this by lowering the risk threshold per title.
