Versioning and the Kindle

You know you’ve been thoroughly indoctrinated into neoclassical economics when you can embrace the benefits of discriminatory pricing and see nothing unfair about different classes of people paying different prices for the same good or more nefariously, companies purposely degrading or adulterating products to sell them at different prices and break consumers into different classes. This process, known as versioning (explained in this paper by economist Hal Varian), explains why many software suites (or Windows Vista) ship standard with certain functions included in the code but disabled, forcing customers to buy the gold or platinum package to have them made operable. (Sometimes this is dubbed crippleware.) It’s also why DVD players don’t play discs from all regions automatically, why more and more services are being stripped from coach class, and so on. If one can see that intentional destruction as actually adding to the sum total of utility in society, then you are really on board.

How is it possible that purposely damaging goods helps a society? The premise, as Varian explains it, is that segments of the market that it would not be profitable to serve at all are put in play by versioning. In general, companies need to create the illusion of premium values among what are essentially the same products in order to get people who are willing to spend more to pay as much as possible. (Branding can play a big role in that procedure, especially with things like corn flakes and bar soaps, cases in which the generic and the branded product are more or less indistinguishable.) The process makes asymmetrical information not the bane of markets, as Akerlof suggested in his “Market for Lemons” paper, but instead a boon of added efficiency for creating profitable price differentials.

What I found disconcerting about Varian’s analysis is his casual aside that “the low-willingness-to-pay consumer always ends up with zero surplus” in the model of versioning he pursues, so they “can safely be ignored in the welfare calculations.” Maybe I am misunderstanding, but doesn’t that mean this model basically ignore the possibility of redistribution of surplus to the “less willing to pay” a.k.a. the poor? “Willingness to pay” seems like a euphemism for “can’t afford to pay because too broke” — a convenient means for evading that particular question of what makes people more or less willing to pay in certain situations, of building in the dubious assumption of consumer sovereignty (the consumer decides what they will do, and the market responds accordingly) directly into the model.

And then then in employing the model itself, Varian sees it as a victory for all of society when producers and rich people (oops, I mean those with a high willingness to pay) accrue more of the social surplus on account of the fact that the total pie has become bigger. When producers make more profit, is that a win for society? If you believe that profit somehow trickles down and around and throughout society, then I guess so. Producers and consumers are basically the same people from a macro perspective. The idea is that more people making and buying more things is inherently good, no matter who is doing the buying and selling. This conflation of social surplus (i.e. profits) with social welfare, though, is the same kind of thinking that would have us ignore growing income inequality in cases when the lower classes’ income growth is not totally stagnant (but just relatively so).

It also helps your conversion to an economistic outlook if you buy into Pareto efficiency as being the basis of ethics, the core method for determining fairness in a society. If no one can be be made better off without any one else having been made worse off as a result, then Pareto efficiency has been achieved. From a non-economist’s view, the problem with this ethical measure is that being better or worse off are relative concepts that are constantly in flux and rely a great deal on the information participants have. I’m not any worse off when I’m flying next to someone who paid have as much as I did for a fare until they tell me about it. In “Markets and Freedoms” Amartya Sen points out that “Pareto efficiency is completely unconcerned with distribution of utilities (or of incomes or anything else), and is quite uninterested in equity.” Most people, however, are interested in those things, and don’t like to see them discarded in favor of output maximization.

Anyway, Varian’s prescription for versioning informational goods reminded me of the Kindle, Amazon’s new electronic book reader that many people seem to be quite taken with. I haven’t used one, but it sounds like an oversize BlackBerry with less functionality. Varian writes: “The first point is the most fundamental: make sure that you design the product so that it can be versioned. That is, the product should be designed in a way that it is easy to reduce its quality in order to sell to a particular market segment.” The fact that the Kindle is WiFi enabled but only allows users to connect to the internet through its proprietary system and access only its sales catalog is an indication of Kindle’s success on that score. Kindle is out to set the terms of e-book reading so that it could implement differential price schemes on the fly: “If the owner of the content controls the browser, then it can choose the features of the browser to enhance the quality of consuming the content…. Controlling the browser allows the seller of content to increase the quality of what it is selling.”

I’m not a long-term skeptic about electronic books — I’m not sentimental about the magic of propping books open and smelling the wood pulp and feeling the grain of the page under my fingers and all that. (I tend to think about what a pain it is to eat and read anything but a magazine at the same time.) But I have a strong inclination to wait until the version of the Kindle comes out that is not locked down, forcing me to pay for information I could get free with WiFi, and giving me only one option for where I can go to replenish my supply of reading material. I’m waiting for the model that is basically a hand-held computer. So I’m hoping the tenets of versioning don’t prevent that device from ever being made.