Under increasing returns, if the number of potential users of a particular good or service is sufficiently small, and the fixed costs are sufficiently large, then the good or service will not be produced even though there exist users whose willingness to pay exceeds the marginal cost of production.On reading this, I am further convinced of my initial reaction, particularly points #2 and #6. This is not a true market failure. Yes, one condition for economic efficiency is that more of a good should be produced if willingness to pay exceeds the marginal cost of production. But that’s only true given that the fixed costs have already been incurred. To justify incurring those fixed costs, it must be true that the consumer and producer surplus (once the fixed costs have been incurred) exceed the fixed costs. To put it another way, fixed costs are actually marginal costs of producing the first unit.
Perhaps the “increasing returns” are supposed to be doing some work I’m not aware of. The term can be used in different ways. If Waldfogel means a situation in which average costs are always decreasing (possibly because marginal costs are falling or constant), that does raise a possibility of market failure – specifically, the kind we usually refer to as “natural monopoly.” But this is not a novel discovery, and in any case, it doesn’t indicate that we have too little product diversity. On the contrary, the natural monopoly is a case is in which it’s efficient to have a single producer (the inefficiency arises from pricing behavior after the monopoly has been achieved). To justify having a second producer with a differentiated product to satisfy minority preferences, the gains to the minority group from having their preferred product must exceed both the fixed costs of production and the added cost for majority consumers who no longer benefit from greater economies of scale.
I haven’t thought the logic all the way through, but if anything I’ll bet there’s an argument that the market could produce too much product diversity from an efficiency perspective. In fact, as I recall, this was the argument made by the theorists of monopolistic competition: that competition in the context of a heterogeneous product leads firms not to fully exploit economies of scale, leading to inefficient excess capacity.
I should emphasize again that since I haven’t read the book, there may be nuances to the argument I’m missing.