EMPIRICAL-ANALYSIS OF A DYNAMIC DUOPOLY MODEL OF COMPETITION

Empirically validating and testing the specification of game theoretic models has received limited attention in the marketing literature. Th e authors provide an econometric framework for estimating the paramete rs of response functions when the observed data in the market place ar e the Nash equilibrium outcomes of an underlying dynamic duopoly game specification. Specifically, the estimation procedure accounts for the joint endogeneity of market shares and marketing efforts of market ri vals using a system of simultaneous equations that included the market response function and the Nash equilibrium conditions. A formal stati stical test is used to detect model misspecification. The empirical an alysis is carried out using data from four product markets: pharmaceut ical, soft drink, beer, and detergent. Comparisons are provided with c onventional estimation of the response function parameters in which th e equilibrium conditions are ignored in the estimation. Managerial imp lications of the empirical results are discussed.