The evidence presented above suggests that Dow's tactics of excess capacity and limit pricing were effective in deterring entry. Yet an alternative explanation is that Dow was simply exercising its technological superiority, derived in part from its huge lead in cumulative production experience. Models of industry evolution with proprietary learning suggest that only a very small number of firms may be viable.17 If the learning curve is steep and one firm develops a very large head start, few firms can coexist, and market shares may be highly skewed. This bears some resemblance to the market structure for magnesium. It is impossible and probably not meaningful to gauge the amount of deterrence stemming from Dow's excess-capacity and limit-pricing policies, as opposed to the firm's cost advantage. The three factors worked together, augmenting the impact that each would have had separately. Dow's pricing strategy reinforced the perception of cost advantage inferred from the data made public following World War II. Similarly, Dow's excess capacity would have been unlikely to deter entry without recognition that this capacity could be mobilized at low cost. And Dow's cost advantage for existing output would have had little deterrent effect without Dow's ability to expand cheaply by improving and reactivating idle sections of its Texas plants. Indeed, once the excess capacity at these plants was exhausted, Dow's behavior shifted dramatically.
The history of the magnesium industry provides rich examples of strategic behavior by a dominant firm. Nevertheless, one must be guarded about drawing generalizations, given the unusual features of the industry and the government's role. Dow's tactics seem consistent with the several deterrence models described above. But Dow's behavior would likely have been very different in the absence of excess capacity built at government expense, and the government's strategic stockpile. These structural features of the magnesium industry are unusual, and their joint influence may well be unique, despite some similarity to other defense- oriented materials (such as aluminum, discussed in Chapter 4 of this volume) during the period following World War II.