Multiscale Adaptive Inference on Conditional Moment Inequalities

TL;DR

This paper introduces a multiscale inference method for conditional moment inequality models, deriving its asymptotic distribution, validating its use through bootstrap, and demonstrating its effectiveness in empirical and simulation studies.

Contribution

It develops a novel multiscale test statistic with an explicit asymptotic distribution and proves its validity, improving inference in models with conditional moment inequalities.

Findings

Test detects local alternatives at optimal rates

Asymptotic distribution is extreme value

Method is adaptive to data smoothness

Abstract

This paper considers inference for conditional moment inequality models using a multiscale statistic. We derive the asymptotic distribution of this test statistic and use the result to propose feasible critical values that have a simple analytic formula, and to prove the asymptotic validity of a modified bootstrap procedure. The asymptotic distribution is extreme value, and the proof uses new techniques to overcome several technical obstacles. The test detects local alternatives that approach the identified set at the best rate among available tests in a broad class of models, and is adaptive to the smoothness properties of the data generating process. Our results also have implications for the use of moment selection procedures in this setting. We provide a monte carlo study and an empirical illustration to inference in a regression model with endogenously censored and missing data.