It's never too LATE: A new look at local average treatment effects with or without defers

Research output: Working paperResearchpeer-review

Abstract

In heterogeneous treatment effect models with endogeneity, identification of the LATE typically relies on the availability of an exogenous instrument monotonically related to treatment participation. First, we demonstrate that a strictly weaker local monotonicity condition - invoked for specific potential outcome values rather than globally -identifies the LATEs on compliers and defiers. Second, we show that our identification results apply to subsets of compliers and defiers when imposing an even weaker local compliers-defiers assumption that allows for both types at any potential outcome value. We propose estimators that are potentially more efficient than 2SLS, even in cases where 2SLS is consistent. Finally, we provide an empirical application to estimating returns to education using the quarter of birth instrument.
Original languageEnglish
Publication statusSubmitted - 20. Jun 2018

Fingerprint

Average treatment effect
Potential outcomes
Participation
Returns to education
Heterogeneous treatment effects
Estimator
Endogeneity
Monotonicity

Cite this

@techreport{31c6c38d22194115946bcf89a0faefcf,
title = "It's never too LATE: A new look at local average treatment effects with or without defers",
abstract = "In heterogeneous treatment effect models with endogeneity, identification of the LATE typically relies on the availability of an exogenous instrument monotonically related to treatment participation. First, we demonstrate that a strictly weaker local monotonicity condition - invoked for specific potential outcome values rather than globally -identifies the LATEs on compliers and defiers. Second, we show that our identification results apply to subsets of compliers and defiers when imposing an even weaker local compliers-defiers assumption that allows for both types at any potential outcome value. We propose estimators that are potentially more efficient than 2SLS, even in cases where 2SLS is consistent. Finally, we provide an empirical application to estimating returns to education using the quarter of birth instrument.",
author = "Dahl, {Christian M.} and Martin Huber and Giovanni Mellace",
year = "2018",
month = "6",
day = "20",
language = "English",
type = "WorkingPaper",

}

TY - UNPB

T1 - It's never too LATE: A new look at local average treatment effects with or without defers

AU - Dahl, Christian M.

AU - Huber, Martin

AU - Mellace, Giovanni

PY - 2018/6/20

Y1 - 2018/6/20

N2 - In heterogeneous treatment effect models with endogeneity, identification of the LATE typically relies on the availability of an exogenous instrument monotonically related to treatment participation. First, we demonstrate that a strictly weaker local monotonicity condition - invoked for specific potential outcome values rather than globally -identifies the LATEs on compliers and defiers. Second, we show that our identification results apply to subsets of compliers and defiers when imposing an even weaker local compliers-defiers assumption that allows for both types at any potential outcome value. We propose estimators that are potentially more efficient than 2SLS, even in cases where 2SLS is consistent. Finally, we provide an empirical application to estimating returns to education using the quarter of birth instrument.

AB - In heterogeneous treatment effect models with endogeneity, identification of the LATE typically relies on the availability of an exogenous instrument monotonically related to treatment participation. First, we demonstrate that a strictly weaker local monotonicity condition - invoked for specific potential outcome values rather than globally -identifies the LATEs on compliers and defiers. Second, we show that our identification results apply to subsets of compliers and defiers when imposing an even weaker local compliers-defiers assumption that allows for both types at any potential outcome value. We propose estimators that are potentially more efficient than 2SLS, even in cases where 2SLS is consistent. Finally, we provide an empirical application to estimating returns to education using the quarter of birth instrument.

M3 - Working paper

BT - It's never too LATE: A new look at local average treatment effects with or without defers

ER -