16 July 2014

AmazonFail 5.1

Still not ready to do internet link sausages — whether it's PTSD from the entirely-predictable end-of-Term nonsense inflicted on us by the Supreme Court, or the side effects of the pig shortage, is rather beside the point — but I do have one more thing to say about the Amazon v. CommercialPublisheroftheWeek (currently Hachette) pricing and availability dispute.

You're all getting it wrong. Just like there is no universally correct solution to the problem of determining greatness in the arts (for example — and this resulted in a great deal of incredulity in the halls of academy, since as a Yank my hero-worship is rather presumed — I find both Moby-Dick and virtually all of Hemingway so fundamentally flawed that they're unreadable), there is no universally correct solution to the problem of determining appropriate pricing and/or distribution methods of personal copies of works in the arts. And the less said about extending the tip of the tail/tale of fiction to the mangy, flea-ridden, undernourished dog of all of publishing, the better.

On the one hand — from a strictly efficiency-is-good economic point of view — having a uniform structure makes sense. Too, a uniform structure and paradigm also reduces (or at least cabins) opportunities for outright fraud.1 But "uniform structure" does not mean "uniform terms," especially in a day of easy-to-manipulate parameters in computerized accounting systems. Just because e-books might be sold to endusers2 does not mean that every e-book needs to be, or should be, on the same terms. Hell, even the most rigid proponents of pricing admit that sometimes there are "special sales," and sometimes Novel A will have a different list price/selling price than Reference Book B (or even Novel C or Short Story D)! Maybe — just maybe — we should admit that more than just the selling price varies for noncomparable works.

The entire argument works as an argument (not even as a lemma, let alone a demonstrable theorem) if, and only if, one assumes that the characteristics of Amazon's market exactly match the corresponding characteristics of CommercialPUblisheroftheWeek's offerings into that market. Any variance at all between them indicates that a rigid, uniform method is inappropriate... except, that is, if one buys the Coase Theorem in its entirety: Not just the assumption of zero transaction costs, but the neglect of translation costs (the cost of converting one form of capital, such as intellectual property, into a different form, such as copies of books) and efficiency as a normative (and not merely descriptive) value. And the less said about the collective-action problem, the better.3

In short, this is the wrong argument. It has little or nothing to do with the future of e-books. It has everything to do with the time-sliced financial reporting (not even fundamental economics) of existing distribution methods. And for that reason, both Amazon and CommercialPublisheroftheWeek should lose... because a victory for either is a loss for both their suppliers and their customers. It's just game theory and the math of stochastic processes, guys (even if HTML doesn't lend itself well to "showing my work").


  1. If you think that the concept of "net receipts," with its rather vague definition, as found in current publishers' contracts concerning e-books is never going to lead to fraud, I could point you to several active disputes regarding precisely that. "Net" is in the eye of the beholder... or stockholder.... Plus, that would be going against three centuries of tradition regarding royalty statements!
  2. And it's a sale. Even if Vernor and similar cases had been correctly decided, and even if the terms of transfer were enforceable — neither is correct — those are for tools directly used in the creation of other works and material, not for static content. If nothing else, an e-book sale (or, for that matter, an MP3 sale) is a sale to the end-user, notwithstanding the economic-terms fight between musicians and music labels (which is about contract language, not statutory interpretation).
  3. I should also note that Professor Solum's numbers in his railroad hypothetical don't work correctly; the equilibrium point for the farmer's contribution is $25, not $50, because at $25 both parties will be avoiding $75 in potential costs. But that's for another time; sadly, his illustration of the Coase Theorem is exactly how neoconservatives and tort-reformers think about the problem...