Production Function Identification Under Imperfect Competition

The presence of imperfect competition introduces distinct challenges when identifying, and estimating, production functions. We start by highlighting that some existing approaches to production function estimation cannot completely abstract away from the presence of imperfect competition in the product market. We then extend these existing approaches to accommodate some additional oligopoly models commonly used in empirical work by using a sufficient statistic approach, and show that the presence of such strategic interactions has important benefits in that they introduces additional exogenous variation that can help identify production functions. We study how to optimally leverage this exogenous variation, both with and without direct data on a firm’s competitors, and use Monte-Carlo experiments to 1) verify that the existence of strategic interactions can identify production functions that would not otherwise be identified, and 2) assess the extent to which what applied researchers observe about competition affects the precision of estimates based on this variation.