Discrete Choice under Risk with Limited Consideration, with Levon Barseghyan and Francesca Molinari. Forthcoming at American Economic Review.
This paper is concerned with learning decision makers' preferences using data on observed choices from a finite set of risky alternatives. We propose a discrete choice model with unobserved heterogeneity in consideration sets and in standard risk aversion. We obtain sufficient conditions for the model's semi-nonparametric point identification, including in cases where consideration depends on preferences and on some of the exogenous variables. Our method yields an estimator that is easy to compute and is applicable in markets with large choice sets. We illustrate its properties using a dataset on property insurance purchases.
I obtain informative bounds on network statistics in a partially observed network whose formation I explicitly model. Partially observed networks are commonplace due to, for example, partial sampling or incomplete responses in surveys. Network statistics (e.g., centrality measures) are not point identified when the network is partially observed. Worst-case bounds on network statistics can be obtained by letting all missing links take values zero and one. I dramatically improve on the worst-case bounds by specifying a structural model for network formation. An important feature of the model is that I allow for positive externalities in the network-formation process. The network-formation model and network statistics are set identified due to multiplicity of equilibria. I provide a computationally tractable outer approximation of the joint identified region for preferences determining network-formation processes and network statistics. In a simulation study on Katz-Bonacich centrality, I find that worst-case bounds that do not use the network formation model are 44 times wider than the bounds I obtain from my procedure.
We present the calibrated-projection MATLAB package implementing the method to construct confidence intervals proposed by Kaido, Molinari and Stoye (2017). This manual provides details on how to use the package for inference on projections of partially identified parameters. It also explains how to use the MATLAB functions we developed to compute confidence intervals on solutions of nonlinear optimization problems with estimated constraints.
Supporting code can be found here. This program is distributed in the hope that it will be useful, but with any warranty; without even the implied warranty of merchantability or fitness for a particular purpose. If you use the software, we ask that you please cite Kaido, Molinari and Stoye (Econometrica, 2019) as the source of the theoretical results and of the code.
Copyright © 2019 Matthew Thirkettle - All Rights Reserved.