Assessing Sensitivity to IV Exclusion and Exogeneity without First Stage Monotonicity
Paul Diegert, Matthew A. Masten, Alexandre Poirier
TLDR
This paper introduces new sensitivity analyses for instrumental variable assumptions, accommodating heterogeneity and relaxing first-stage monotonicity.
Key contributions
- Develops new sensitivity analyses for instrumental variable (IV) exclusion and exogeneity.
- Accommodates arbitrary heterogeneity in treatment effects and relaxes first-stage monotonicity.
- Derives identified sets for potential outcomes and average treatment effects under nonparametric relaxations.
- Provides computationally tractable methods for estimating these identified sets.
Why it matters
This paper addresses critical limitations in instrumental variable analyses by providing robust methods to assess assumption sensitivity. It allows researchers to draw more reliable causal inferences even when standard IV assumptions are violated.
Original Abstract
Exclusion and exogeneity are core assumptions in instrumental variable (IV) analyses, but their empirical validity is often debated. This paper develops new sensitivity analyses for these assumptions. Our results accommodate arbitrary heterogeneity in treatment effects and do not impose any monotonicity requirements on the first stage. Specifically, we derive identified sets for the marginal distributions of potential outcomes and their functionals, like average treatment effects, under a broad class of nonparametric relaxations of the exclusion and exogeneity assumptions. These identified sets are characterized as solutions to linear programs and have desirable theoretical properties. We explain how to estimate these solutions using computationally tractable methods even when the linear program is infinite-dimensional. We illustrate these methods with an empirical application to peer effects in movie viewership, using weather as a potentially imperfect instrument.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.