Interpreting Instrumental Variable Estimands with Unobserved Treatment Heterogeneity: The Effects of College Education

TL;DR

This paper clarifies how instrumental variable estimands relate to unobserved treatment heterogeneity, develops a method to incorporate non-monotonic instruments, and applies it to estimate college education returns, revealing significant wage effects.

Contribution

It introduces a new framework for interpreting IV estimands with unobserved treatment heterogeneity and develops a method to include non-monotonic instruments in treatment effect estimation.

Findings

IV estimands are weighted averages of unobserved treatment effects.

Differences in IVs can cause estimand variation even without heterogeneity.

Estimated lifetime wage returns to college range from 7% to 30%.

Abstract

Many treatment variables used in empirical applications nest multiple unobserved versions of a treatment. I show that instrumental variable (IV) estimands for the effect of a composite treatment are IV-specific weighted averages of effects of unobserved component treatments. Differences between IVs in unobserved component compliance produce differences in IV estimands even without treatment effect heterogeneity. I describe a monotonicity condition under which IV estimands are positively-weighted averages of unobserved component treatment effects. Next, I develop a method that allows instruments that violate this condition to contribute to estimation of treatment effects by allowing them to place nonconvex, outcome-invariant weights on unobserved component treatments across multiple outcomes. Finally, I apply the method to estimate returns to college, finding wage returns that range from 7\% to 30\% over the life cycle. My findings emphasize the importance of leveraging instrumental variables that do not shift individuals between versions of treatment, as well as the importance of policies that encourage students to attend "high-return college" in addition to those that encourage "high-return students" to attend college.