Robust inference in deconvolution

Kotlarski's identity has been widely used in applied economic research based on repeated‐measurement or panel models with latent variables. However, how to conduct inference for these an open question two decades. This paper addresses this problem by constructing a novel confidence band the density function of variable repeated measurement error model. The builds our finding that we can rewrite as system linear moment restrictions. Our approach is robust do not require completeness. controls asymptotic size uniformly over class data generating processes, and it consistent against all fixed alternatives. Simulation studies support theoretical results.