Tom Lehrer understated the complexity of understanding publishers' royalty statements. It's more than just the possibility that some of the numbers are in base 8 (which tracks the typical understatement of royalties fairly well, leading perhaps to more questions than it answers!). Instead, there are other problems that only doing the math will allow one to see—and the math necessary is far from self-evident.

Consider, for example, the trade fiction casebound (hardback) publishing agreement. For a midlist author, the royalty basis will boil down to something like this, if the author's agent has negotiated better than would my dog:

10% of the cover price on the first 5,000 copies, then

12% of the cover price on the next 5,000 copies, then

15% of the cover price for each copy thereafter,

for copies sold in the ordinary course of business

Time for the math lesson, keeping in mind the objections raised here two days ago to a foreseeable trend that already has precedent in electronic rights: pegging the royalty to the "publisher's net proceeds." We'll leave aside for the moment the problems with defining "net proceeds" and only shudder at the possibility that Hollywood's fifty-page definition might be imposed (and also leave aside the antitrust implications). For the moment only, though.

From an author's perspective, alteration of the "standard" contract terms needs to maintain author income (in reality, author credits toward advances, since most books in this class do not earn out their advances) at the current level. Since the vast majority of fiction sales are at a "long discount"—currently 55%, but there is pressure from some retailers to raise it to 60%—equivalent income-producing terms, assuming for simplicity that the "net proceeds" are the amount the publisher receives for each book at long discount, would look something like this:

10%/(net) = 10%/45% = 22.22% of the net proceeds on the first 5,000 copies, then

12%/(net) = 26.67% of the net proceeds on the next 5,000 copies, then

15%/(net) = 33.33% of the net proceeds for each copy thereafter,

for copies sold at long discount

Things get even more interesting if the "long discount" is set at 60%:

10%/(net) = 10%/40% = 25% of the net proceeds on the first 5,000 copies, then

12%/(net) = 30% of the net proceeds on the next 5,000 copies, then

15%/(net) = 37.5% of the net proceeds for each copy thereafter,

for copies sold at long discount

It should surprise nobody that publishers have been imposing nonnegotiable alternative contracts and modifications to existing contracts that look substantially less favorable to the authors. This is more typical of what I have seen:

20% of the net proceeds on the first 5,000 copies, then

24% of the net proceeds on the next 5,000 copies, then

30% of the net proceeds for each copy thereafter,

for copies sold in the ordinary course of business

On average, this represents an income loss to the authors of well over 10%. The publishers typically claim that short-discount and direct sales justify the lower per-copy percentage, but sales figures for a typical work of fiction or serious trade nonfiction do not bear that out. Considering the accounting efficiency that can be had by no longer working on a "number of copies" basis for returns, that is an even greater savings for the publisher. Since the base royalty rates for trade paperbacks and mass-market paperbacks tend to be quite a bit lower than those for casebounds, there is even more opportunity for shenanigans in those contracts.

Yes, this is going somewhere. Probably to hell in a handbasket, but we'll just have to see.