Identification and estimation of partially identified models with with applications to industrial organization

Open Access
Li, Xian
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
May 29, 2018
Committee Members:
  • Andres Aradillas-Lopez, Dissertation Advisor
  • Andres Aradillas-Lopez, Committee Chair
  • Patrik Guggenberger, Committee Member
  • Coenraad Arnout P Pinkse, Committee Member
  • John Richard Howell, Outside Member
  • Paritial Identification
  • Demand Estimation
The thesis includes three chapters. The first chapter studies a type of partially identified models defined by moment equalities with latent variables and proposes a novel method to identify and estimate the parameters of interest. The second chapter applies the method in chapter one to demand estimation when the observed market shares cannot correctly represent the true market shares. The third chapter studies price competition models when firms are facing capacity constraints, and provides methods to partially or point identify and estimate the parameters of interests in both demand and supply function.