Universal Inference for Incomplete Discrete Choice Models

TL;DR

This paper introduces a robust, finite-sample valid inference method for set-valued predictions in incomplete discrete choice models, enabling reliable confidence intervals without resampling or tuning.

Contribution

It develops a universal inference-based procedure that handles model incompleteness and nuisance parameters, avoiding resampling and tuning parameters.

Findings

Provides a tractable inference method with finite-sample validity.

Enables construction of confidence intervals for counterfactuals.

Applicable to models with both discrete and continuous covariates.

Abstract

A growing number of empirical models exhibit set-valued predictions. This paper develops a tractable inference method with finite-sample validity for such models. The proposed procedure uses a robust version of the universal inference framework by Wasserman et al. (2020) and avoids using moment selection tuning parameters, resampling, or simulations. The method is designed for constructing confidence intervals for counterfactual objects and other functionals of the underlying parameter. It can be used in applications that involve model incompleteness, discrete and continuous covariates, and parameters containing nuisance components.