| In two-sided markets with positive indirect network effect, the number of consumers on a platform depends on the entry and prices of the affiliated products, and the success of the affiliated products depends on the number of consumers. Moreover, platform markets are often inherently dynamic environments due to the durability of platform intermediaries and the affiliated products. In the dynamic two-sided market environment, overpricing one side of the market not only discourages demand on that side but also discourages participation on the other side, and so can lead to a death spiral.;This dissertation presents a dynamic structural model to analyze consumers' purchase decisions for competing hardware platforms and their affiliated software products, and software firms' dynamic pricing and entry decisions. Consumers are heterogeneous and forward-looking and have rational expectations about software entry and prices in the future. They choose when to purchase the hardware and the affiliated software. Software firms decide on price and whether to affiliate with a platform, accounting for competitors' reactions, consumer composition, and consumer forward-looking behavior.;This dissertation provides a new practical methodology for structural estimation of dynamic equilibrium models. It combines the Bayesian Markov Chain Monte Carlo (MCMC) algorithm and the dynamic equilibrium model solution algorithm into a single algorithm that estimates the parameters and solves the model simultaneously. It partially solves for each agent's best response and value function given that other agents play the equilibrium strategies at each draw of parameter vectors. It uses those pseudo-best response functions and pseudo-value functions to non-parametrically approximate the equilibrium strategies and the value functions at the current trial parameter vector.;The estimation method is applied to the U.S. fifth-generation video game industry (May 1995 - February 2002). Counterfactual experiments suggest that the installed base of users and the software variety would significantly increase if a platform had reduced its console price, or had subsidized its software entry, or had lowered its royalty fee on independent software producers.;JEL Code: C11, D12, D22, D43, L11, L86.;Keywords: Bayesian Estimation, Dynamic Pricing, Dynamic Game, Two-Sided Market, Video Game Industry. |
