Dice Question Streamline Icon: https://streamlinehq.com

Identifying compliers in LATE settings

Determine which observational units are compliers—that is, those whose treatment status is affected by the instrumental variable—in applications that estimate local average treatment effects with a binary instrumental variable and a binary endogenous regressor, to enhance interpretation and external validity of instrumental-variables estimates.

Information Square Streamline Icon: https://streamlinehq.com

Background

Under monotonicity and related assumptions, two-stage least squares identifies a Local Average Treatment Effect (LATE) for the subset of units whose treatment is shifted by the instrument (the compliers). However, applied researchers typically cannot observe who belongs to this subpopulation, limiting the interpretability and policy relevance of LATE.

Making compliers identifiable—or at least characterizing them systematically—would clarify for whom the estimated effect applies and would support more informed extrapolation or targeting in agricultural and applied economics contexts.

References

In the case of a binary instrumental variable and a binary endogenous explanatory variable, the LATE indicates the average treatment effect on those subjects that ‘comply’ with the instrumental variable, while the effects on the ‘always takers’ and the ‘never takers’ remain unidentified and in most cases it is unknown who the ‘compliers’ actually are.

— Estimating Causal Effects with Observational Data: Guidelines for Agricultural and Applied Economists (2508.02310 - Henningsen et al., 4 Aug 2025) in Section 3 (Instrumental-Variable Methods)