Partial Identification of Marginal Treatment Effects with discrete instruments and misreported treatment

TL;DR

This paper derives partial identification results for marginal treatment effects when treatments are misreported and instruments are discrete, using nonparametric assumptions, with an application to food stamps and health.

Contribution

It introduces novel partial identification methods for MTE with misreported binary treatments and discrete instruments, expanding the scope of causal inference under data imperfections.

Findings

Identification of MTE under misreporting and discrete instruments

Application to food stamps and health outcomes

Framework for nonparametric partial identification

Abstract

This paper provides partial identification results for the marginal treatment effect ($MTE$) when the binary treatment variable is potentially misreported and the instrumental variable is discrete. Identification results are derived under different sets of nonparametric assumptions. The identification results are illustrated in identifying the marginal treatment effects of food stamps on health.