Tolerating defiance? Local average treatment effects without monotonicity

By C. Chaisemartin. 2017

109 citations recorded by the source corpus.

We know that instrumental variable (IV) estimates a causal effect if the instrument satisfies a monotonicity condition. When this condition is not satisfied, we only know that IV estimates the difference between the effect of the treatment in two groups. This difference could be a very misleading measure of the treatment effect: it could be negative, even when the effect is positive in both groups. There are a large number of studies in which monotonicity is implausible. One might then question whether we should trust their estimates. I show that IV estimates a causal effect under a much weaker condition than monotonicity. I outline three criteria applied researchers can use to assess whether this condition is applicable in their studies. When this weaker condition is applicable, they can credibly interpret their estimates as causal effects. When it is not, they should interpret their results with caution.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/d491d2ab031d117ae57938e928044803d463661f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Judging Judge Fixed Effects (2019, background): “In this framework the result of Theorem 2 is even stronger: E [Yi|Ji = j] is not only continuous, but a linear function of propensities. de Chaisemartin (2017) presents a weaker monotonicity assumption under which a proper weighted average of treatment effects is still identified.”