This started in response to James D. Macdonald’s true but less than explicit statement over at Making Light:
I see that faux-statistic that 70% of books don't earn out, therefore 70% of books don't make a profit all over the place, and it's equally false all over the place. Publishers start making a profit on their books long before that book earns out.
But my explanation got rather long, so I brought it over here, rather than going on and on and on in the comments to Teresa’s post.*
It seems logical that if a book doesn't start earning the author royalties above the advance that the publisher paid, it must not have been a profitable endeavour for the publisher. At least, this seems logical until you consider the costs of making a book.
First, some of what we in the textbook publishing world call Key Terms, and list at the beginning of every chapter:
Royalties are a percentage of the sale price of the book that the author earns for each copy sold. So if a book has a sale price of $20, and the author recieves 15% of that, then for every copy of the book sold, the author will receive $3.
The author's advance is a sum that the publisher pays up front, before the book has been published. It's kind of a bet that the publisher makes that the book will earn more than that—the advance goes against the royalties. So if the author of our hypothetical $20 book was paid a $10,000 advance, he or she wouldn't see a royalty cheque until more than 3,334 copies had sold.
When our hypothetical book sells its 3,334th copy it has earned out the advance the publisher paid to the author.
Most authors’ contracts are more complicated than my example—I chose round numbers because I’m lazy and didn’t want to do too much arithmetic, and didn’t mention anything about sales of subsidiary rights because I wanted to keep the example straightforward—and most books don’t sell for $20, but you can still see that the author’s advance is a cost the publisher pays in order to get a manuscript, and that the amount that the author can earn from his or her work is advance + royalties after the book earns out its advance.
The author’s advance represents only a tiny fraction of the costs the publisher pays in order publish a book. If we just paid the author an advance and never spent any more money on a book, we’d have a manuscript, which may or may not be remotely publishable. In other words, we’d have a Word or .rtf file (unformatted, or possibly weirdly formatted), or maybe (if we’re a really old-fashioned publisher, or we have a much-beloved really old-fashioned author) and possibly some sketches or blurry copies of photos. In order to make a book out of these files or pages, the publisher is going to have to pay some people to work some magic.
These costs fall into two categories**: recurring costs and non-recurring costs.
Paper, print, and binding (PPB) represent the other major recurring cost centres. We have to pay for every copy of a book we print. As the number of books we print goes up, the cost to print each book goes down, because of economies of scale.
Editing (substantive/line editing, copy editing, and proofreading), design (interior and cover), formatting, permissions, and illustrations constitute some of the major non-recurring costs on an edition of a book—once we've paid a copy editor, we can print as many copies as we think will sell for a given book.
Then there are costs for marketing and sales, promotions, publisher's overhead (paying for premises, warehousing, staff salaries—which, I feel compelled to point out, are not extravagant—equipment, and all the other things the publisher needs in order to keep making books), keeping a lawyer on retainer, and probably some stuff I’m forgetting.
When deciding whether to publish a book, how much to spend on said book, and how much to charge the reader (and the booksellers, who get a discount) for the book, the publisher prepares a Profit & Loss statement (a P&L) (see the redoubtable Anna Louise Genoese's breakdown of how this works for trade books) to figure out whether the book is likely to make a profit given what the publisher believes the sales are likely to be (we use a variety of divinatory methods to predicts sales, including looking at sales of the author’s previous books, acquiring advance sales of the book, guessing at the number of potential buyers—in textbook publishing this includes all the schools that offer the course for which a textbook is being created; in trade it’s more nebulous than that—and factoring any external factors that might drive people to buy that book, such as the book’s being about a current hot topic, or the author doing something scandalous that might generate interest). The publisher calculates all the non-recurring costs, then estimates the recurring costs based on how many copies it thinks it should print (based on what the divinations have revealed that sales are likely to be). Overhead or operational costs don’t get factored into the P&L; however, it should be clear that publishers use the profits from their sales to pay their overhead, just as most sales-based businesses do. This is why we want books to make a decent profit: If they don’t, then nobody gets paid, and our computers die and we have to make them work with duct tape and paperclips, and we have to edit in the dark and cold with aged computer running obsolete software, and it’s just miserable. Publishing is a Business.
So, no matter how much an editor loves a book, no matter how badly an editor thinks that the world needs this book, if the editor doesn’t think that they can make it a profitable book, they’re probably not going to publish it. Because we have to pay the author’s advance, and the fees for all the costs associated with making the book and getting it to market, and we’d really prefer to have money left over to pay the staff (like me!), and maybe change the lightbulbs and upgrade some software.
So, as any business does before setting out on a new venture, we look at the probable costs, the likely profits, do what we can to minimize the former and maximize the latter, and hope that good things happen and bad things don’t.
* About Nicholas Borelli, to whom Jim is responding, and his blog post positing a feminist conspiracy that keeps his writing from being published I will say little, right now. I’ve said most of what I need to say about the Nicholas Borellis of the world.
** Three, if you consider overhead a cost, but people generally don’t. Overhead represents several different cost centres, but they float over all the income-generating projects, so don’t get addressed in project-specific budgeting or P&Ls.