No, Virginia, e-books don’t publish themselves

June 18, 2012, 4:00 PM UTC

Image via The New Yorker

FORTUNE — Whenever I write about the e-book business — especially in the context of the Justice Department’s antitrust suit against Apple (AAPL) and five book publishers — someone in the comment stream invariably suggests that e-books are vastly overpriced because the publishers who sell them incur none of the usual costs of printing and distribution.

E-books don’t write or edit themselves, of course, but I’ve never seen a good accounting of what it actually costs to publish one. So I was gratified to find, high up in “Paper Trail,” Ken Auletta’s smart piece about the e-book business in the current issue of the New Yorker, this paragraph:

“E-books are cheaper to produce, by about twenty per cent per book, because they do away with the cost of paper, printing, shipping, and warehousing. They also eliminate returns of unsold books—a significant expense, since thirty to fifty per cent of books are returned. But they create additional costs: maintaining computer servers, monitoring piracy, digitizing old books. And publishers have to pay authors and editors, as well as rent and administrative overhead, not to mention the costs of printing, distributing, and warehousing bound books, which continue to account for the large majority of their sales.

Twenty percent doesn’t seem like much of a savings, especially considering that Amazon (AMZN) has been selling electronic versions of books that cost $30 in hardback for anywhere from $14.99 to $4.99 to $0.00

The accounting gets even more complicated when you consider that most books cost publishers more than they earn. Amazon’s Jeff Bezos argues that traditional publishing is infuriatingly wasteful, and that his model is far more efficient and better for both readers (by lowering prices) and would-be authors (by eliminating the “gatekeepers” that prevent them from reaching readers).

As Auletta points out …

The traditional model has advantages for authors, though, particularly in publishers’ function as venture-capital firms. When an author sells a book proposal to a publisher, he receives an advance against royalties, which helps underwrite research and writing. Most of these advances are never earned back. But books that sell well support the ones that don’t. In a good year, this earns the publishers a modest profit, and it allows more authors to take risks in starting new projects—which is to say, it supports a class of professional writers. In a more efficient world, publishers would pay authors who write best-selling books and rarely pay those who don’t, an alarming prospect for most serious readers and writers. According to a recent survey by the Web site Taleist, half of all self-published authors make less than five hundred dollars a year from their writing. If publishers don’t have the money to pay advances, Young says, “we’re going to have fewer books of quality. The impact on authors could be huge.”

As someone who has worked in print and electronic publishing, I thought Auletta’s piece offered a pretty even-handed account of both sides of the e-book wars. If you have a subscription to the New Yorker, it’s available as a pdf here.