Andrew Odlyzko, author of such gems as Content Is Not King, has a new paper available: Privacy, economics, and price discrimination on the Internet.

Perfect price discrimination has long been raised as one of the justifications for DRM (price discrimination depends on preventing arbitrage, that prevention may be enforced by DRM-backed no-resale clauses); Odlyzko suggests that consumers tend to rebel against overt price discrimination, and will therefore be subject less to DRM than to more covert forms such as bundling. I'm not sure that reduces the dependence on DRM, since DRM and anti-reverse engineering law often enforce the bundling. Price discrimination is one explanation for Lexmark's strategy: Selling printers at a loss but making it up on toner cartridges enables Lexmark to charge use-based pricing on the package.

I think we'll also see trusted computing called into the service of perfect price discrimination. With trusted computing, everyone may be able to get (only) a customized version of software or media, at a "custom" price if a vendor chooses.

The non-monetary privacy costs are high, however. Price discrimination demands the ultimate "know your customer." We trade personal data for frequent flyer discounts, but also for an identification that may allow sellers to charge us more when they recognize we need a product or can afford to spend more. The more price discrimination becomes part of the fabric of online transactions, the less we'll economically be able to opt-out of identification schemes.