Working Paper: NBER ID: w21999
Authors: Ariel Pakes
Abstract: This paper investigates progress in the development of models capable of empirically analyzing the evolution of industries. It starts with a parallel between the development of empirical frameworks for static and dynamic analysis of industries: both adapted their frameworks from models taken from economic theory. The dynamic framework has had its successes: it led to developments that have enabled us to control for dynamic phenomena in static empirical models and to useful computational theory. However when important characteristics of industries were integrated into that framework it generated complexities which both hindered empirical work on dynamics per se, and made it unrealistic as a model of agent behavior. This paper suggests a simpler alternative paradigm, one which need not maintain all the traditional theoretical restrictions, but does maintain the core theoretical idea of optimizing subject to an information set. It then discusses estimation, computation, and an example within that paradigm.
Keywords: No keywords provided
JEL Codes: L13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| inclusion of important industry characteristics (L60) | complexity (C60) |
| complexity (C60) | feasibility of empirical analysis (C12) |
| computational advancements (C63) | ability to analyze dynamic models (C69) |
| computational advancements (C63) | effectiveness of empirical analysis (C52) |