This paper is in three parts about the market factor in contingency theory: (1) We focus on the dual structure of markets; the internal structure of relations among producers vs. the external structure of buying and selling with other markets. We use a network model to describe the association between performance and the dual structure of American markets from 1963 to 1992. (2) We reverse-engineer the network model to infer the "effective" level of competition among producers in each market. Effective competition, a measure of competitve intensity, is inferred from observed market profits predicted by the market network of dependence on other sectors of the economy. Producers with profit margins higher than expected from observed market structure must face an "effective" level of competition lower than the level implied by the observed structure. Instead of predicting peformance from internal and external market structure, we use data on performance and external structre (the more reliable and detailed data) to infer internal structure. (3) We demonstrate the research value of the effective competition variable for its relaiability (illustrated by automatic adjustment for the exogenous shock of imports in 1982), its accuracy (illustrated by revealing the contingent value of a strong corporate culture in Kotter & Heskett's, 1992, study), and as a market factor integrating case with comparative research. We close discussing the market conditions measured by effective competition, which, as an unobserved variable, is more subject than observed variables to misinterpretation.