by Cass Sunstein

Chris Anderson's new book, The Long Tail, is causing a considerable stir. To make a complex story short, Anderson argues that companies can, and do, make increasing amounts of money by catering to niche markets through a large volume of products (books à la Amazon.com, movies à la Netflix), many of which are bought by very few people. At a bookstore, very little money is made from the poor sellers (at the long tail of the distribution). At Amazon.com, by contrast, the immense stock of books, and the large customer base, can ensure that significant aggregate sales come from the long tail. Anderson sees this as an important trend. With the aid of the Internet, and of other modern technologies, it is often nearly costless to sell not just the blockbusters, but also goods that cater to small markets; the total profits from doing so may be very high. "Niche" is a key word in Anderson's book. Anderson is a terrific writer, and his book is a lot of fun, and most of his crucial points are right. What is also remarkable is his lack of self-consciousness about what might be wrong with a world of niches. He writes as if the power to choose the particular good that each particular person particularly wants is an unambiguous good--as if there is little to do but to notice and celebrate this process. I do not know whether Anderson is a follower of Milton Friedman, but one of the most striking features of the book is its occasionally breathless quality, as if freedom and the good life are promoted by, and maybe even captured in, the opportunity to choose what is specifically sought on either the large head or the long tail. The mixture of justified enthusiasm for amazing technological innovations and almost-never-even-slightly-questioned libertarianism is striking. I believe that this mixture captures a great deal of thinking about the Internet, even among its most creative and sharpest analysts. What Anderson does not seriously engage is the risk that "niches" will have adverse effects on aspects of the shared culture and also promote fragmentation, particularly along political lines. (In previous posts--here and here I have discussed ideological amplification, in which like-minded people, speaking largely with one another, go to extremes; "niches" facilitate this process.) He does offer a brisk discussion of the problem of fragmentation, but he rests content with general observations about how many people are curious and how niches include shared cultures of different kinds. I hope I'm not being unfair to Anderson. His book is both excellent and illuminating. What's striking is he appears not to appreciate how the world of niches--even if on balance a big improvement over what preceded it--might have a downside in the domains of both politics and culture.