Katz and Smith are, of course, correct in talking about the importance of production flexibility in the industrial organization literature. But there is also a specific example of this, not particularly relevant to their case but relevant in many other cases, in which a company is so designed that the costs do not vary much over a wide range of output. There seems to be no doubt that many corporations are deliberately designed to provide one of these two kinds of flexibility. They will have either the ability to easily switch to another product or the ability to operate on many different production rates at about the same cost.
It is my purpose to argue that although this may be true in fact, the standard economic explanations are false. It would appear to be a dominated strategy in a competitive industry. I do not argue against the theory in areas where competition is highly restricted.
The argument is fairly simple. If we invest, as Katz and Smith argue, in flexibility in the sense that we design our factories so that they not only can produce the defense component of their example but also something else in the event that the demand for the defense component falls off, we will inevitably have a higher cost for producing the defense component than we would if we had not done so. If that is not true, of course we are not actually investing in the additional flexibility. 1
If we then compare two plants in an industry, one of which has been designed with flexibility; hence, has a somewhat higher minimum costs, and the other, which has been designed to produce one product at a volume giving the lowest possible cost, 2 it is obvious that, in the long run, the specially designed plant will drive the other out of business. With a fluctuating demand, the manager of the specially designed plant simply always runs his plant at its optimal rate and sells the product for whatever he can get.
The profits (losses) per unit of sale will fluctuate depending on business conditions, but they will always be higher (lower) than those of the flexible plants. After all, his costs are at his most efficient point, which gives him lower costs than the flexible plants. In essence, the entire volume fluctuation is imposed on the flexible plants. This is true whether the flexibility takes the form of changing production rate or switching product.