Chris Shannon

ABSTRACT:

This paper studies the testability of theories when data might be subject to measurement error. The paper considers a general revealed preference framework for rationalizing data and refuting theories subject to noisy observations. The paper gives several conditions under which features of a model might be estimated or tested using such data, making use of ideas in topological data analysis. Examples including consumer demand and general equilibrium illustrate the main results.

BIO:

Chris Shannon is a professor at UC Berkeley. She earned undergraduate degrees from the University of Kansas, and a PhD from Stanford University. From 2012-2013, she was the Richard Merkin Professor of Economics and Mathematics at Caltech. She is currently the Richard and Lisa Steiny Professor of Economics and Professor of Mathematics at UC Berkeley.