This article was updated on June 12, 2026.

Books are available in more formats and channels than ever today—a boon to reading everywhere. Readers can read books in print, large print, ebook, audiobook, and serialization formats and can access them through a variety of in-person and online platforms, including through ebook and audiobook subscription services and library apps. This is something to be actively celebrated, given the enormous competition for people’s attention with other media today. Unfortunately, authors’ earnings have not kept pace with the abundant availability of books, but have been in steep decline for the last decade and a half.

There are many factors contributing to this decline in book earnings, as we have discussed in connection with our earnings studies. A recent survey sponsored by the Authors Guild and conducted by the Codex Group, a book industry research and analytics firm, indicates yet another likely cause that surprised us: not as many active readers actually buy the books they read today. Instead, almost two-thirds of readers obtain books for free – whether from friends, personal collections, libraries, pirate sites, or other free sources.

The survey asked regular readers (those who had read at least one book in the last month and four books in the last year) where they obtained the books they read or listened to in the prior month. Only about a third of readers and listeners bought new royalty-generating copies of books or audiobooks, including those acquired through a paid subscription service.

The survey further found that only 19 percent of books in text format were bought as new print or ebook copies (which give authors the highest royalties) while 6 percent were accessed through subscriptions, many of which pay authors a fraction of what they would earn on individual sales. Another 10 percent were bought used, sales which provide authors with no payment. 29 percent of books read were acquired through libraries, and the remaining 35 percent of books read were either borrowed from friends or family, came from a reader’s personal collection, or acquired through another source, such as piracy.

Among digital audiobooks, 36 percent of books listened to in the survey period were either purchased or acquired through a paid subscription, with subscriptions outranking audiobook purchases 2-to-1. 37 percent of audiobooks were sourced through digital library borrowing.

The shift from purchases to low or non-royalty generating channels has profound implications for the writing profession. According to the Authors Guild’s last income survey in 2023, the median income of a full-time author from their books is $10,000 a year, and $20,000 from all writing-related work combined. This is down roughly 42 percent since 2009 when the first consumer Kindles entered the marketplace.

The survey also found that the readers most likely to borrow from a library instead of buying a new book are college educated, employed full-time, and earning more than $75,000 a year—not generally those who cannot afford to buy books. At the same time, active library members read 16 percent more books than non-members and were nearly twice as likely to listen to audiobooks.

For the full survey results, which include data on book reading habits of the respondents by gender and format, as well as source, see here.

Some of the key findings are below.

Key Findings

Where Readers Sourced the Books They Read

Text and Audio Formats Combined

• Only 36 percent of people who read a book or listened to an audiobook in the last month bought a new book or audiobook (or obtained one through a paid subscription service) that generated royalties for the author in that time. Nearly two-thirds purchased nothing new.

Text Formats

• Only 25 percent of books (ebook and print) read in the last month were bought new or obtained through a subscription.
  • Just 19 percent of books in text format (print and ebook) read in the prior month were bought new.
  • 6 percent were obtained through a paid subscription such as Kindle Unlimited (which pays much less to authors).
• 10 percent were bought used.
• 29 percent came from public library borrowing—both print and digital.
• 16 percent were borrowed for free from other sources (including piracy) and another 19 percent were from personal collections.

Audio Formats

• 36 percent of digital audiobooks consumed were either purchased new or accessed through paid subscriptions like Audible. 37 percent of digital audiobooks were borrowed from libraries.
• 27 percent were acquired from other sources (including pirated copies).

Brand-Name Authors Feel the Impact Most

The survey tested a list of 35 major “brand-name” authors—including Stephen King, Colleen Hoover, Rebecca Yarros, James Patterson, and Sarah J. Maas—and found a pronounced borrowing effect:

• Fans of these authors were 35 percent more likely to have sourced their last book read from a library than readers of other authors.    
• They were 28 percent less likely to have purchased the book new.

In other words, the authors with the highest name recognition—and whose books libraries stock most aggressively—see the greatest substitution of borrowing for buying.

Gender Differences in Borrowing and Buying

• Men were more likely to purchase: 42 percent of male past-month readers bought a new book or audiobook or obtained one through a paid subscription, compared with 33 percent of women.
• For text formats (print books and audiobooks), women sourced 25 percent more books from public libraries than from new purchases (30 percent borrowed vs. 24 percent bought new).
• The audiobook gap is even more striking. Women sourced 44 percent of their digital audiobooks from libraries and just 32 percent from paid channels. Men showed the reverse pattern: 44 percent from paid sources and only 20 percent from libraries.

Audiobooks: Libraries and Subscriptions Dominate

• Public libraries were the single largest source of digital audiobooks, representing 37 percent of books consumed.
• Among paid channels, subscriptions (like Audible) accounted for twice the share of individual purchases (24 percent vs. 12 percent).
• Active library members bought 40 percent fewer audiobooks (either purchased outright or through a subscription service) than non-library users.

Libby

• Libby/OverDrive dominated digital library lending, commanding an estimated 70 percent of ebook borrows and 74 percent of digital audiobook borrows.
• Hoopla (14 and 16 percent, respectively) and CloudLibrary (12 and 8 percent, respectively) were distant runners-up. The Palace Project registered at 1 percent for ebooks and 0 percent for audiobooks.

Interesting Data Points Regarding Library Usage

• 59 percent of past-month readers consider themselves current active public library members.
• Active library members are more likely to have household incomes above $75K, hold graduate degrees, and be employed full-time.
• Active library users read 16 percent more books than non-library readers—but bought 42 percent fewer new books.
• Active library members are 14 percent more likely to live in suburban areas.
• Those earning under $75K are 28 percent more likely to not be active library members.
• Peak active library usage occurs among readers aged 25–44.
• Active library members are 46 percent more likely to listen to audiobooks and substantially more likely to read in both print and ebook formats than other readers.
• Those who read an ebook in the last month read at least 50 percent more total book units in the last month than those who didn’t.
• Non-library member ebook readers bought 73 percent more new books in the last month (2.2) than library members did (1.3).
• Non-library member, non-ebook readers also bought 73 percent more new books in the last month (1.1) than library members did (0.7).

View the full report here (PDF).