Justin Wolfers points to the new book, The Tyranny of the Market by , about “the ways in which markets can fail us”. In a recent Slate article, Waldfogel summarizes his argument:

After all, free-market economists have told us for decades that we should rely on market decisions, not the government, to meet our needs, because it’s the market that satisfies everyone’s every desire.

[…]

Two simple conditions that prevail in many markets mean that individual taste alone doesn’t determine individual satisfaction. These conditions are 1) big setup costs and 2) preferences that differ across groups; when they’re present, an individual’s satisfaction is a function of how many people share his or her tastes. In other words, in these cases, markets share some of the objectionable features of government. They give bigger groups more and better options.

Perhaps I am missing something, but, while what Waldfogel discusses could be described as a market disappointment, I don’t think it can be called a “market failure.” Fixed  certainly do stop some markets from emerging, but this is actually efficient; those fixed costs are real, and if we (government) decided to somehow force society to pay those fixed costs in order to serve those markets it would have a real, negative effect on welfare. There are some niche markets for which it doesn’t make economic sense (in a social sense) to serve. If Native Americans greatly valued specially designed Nike shoes, they would have been willing to pay a high premium for them, and Nike probably would have invested in the Native American shoe market earlier. If they are not willing to pay premium for them, it is and indication that it  does not make economic sense (in a non-distributional sense) to serve that market.
Waldfogel’s point seems so obviously wrong, that I feel as though I must have missed something, so perhaps I will read his book.

It is also possible that I simply misunderstand the origin of Waldfogel’s concerns. Perhaps Waldfogel means that people in general have a strong preference for seeing niche markets served. Then there would a positive externality created by serving a niche market and that could certainly be described as a market failure (though not one very high on my list of market failures to fix). This seems like a very contestable empirical question. I certainly haven’t seen any data, but I doubt this feeling is type of preference is widespread. It is also possible that Waldfogel is concerned about the distributional effect of these fixed cost, perhaps more poor people have out-of-the-mainstream needs and desire, but I don’t think this is what he is talking about.