RT Generic
T1 Treatment effect identification using alternative parallel assumptions
A1 Mora, Ricardo
A1 Reggio, Iliana
A2 Universidad Carlos III de Madrid. Departamento de EconomÃa,
AB The core assumption to identify the treatment effect in difference-in-differences estimators is theso-called Parallel Paths assumption, namely that the average change in outcome for the treated inthe absence of treatment equals the average change in outcome for the non-treated. We define afamily of alternative Parallel assumptions and show for a number of frequently used empiricalspecifications which parameters of the model identify the treatment effect under the alternativeParallel assumptions. We further propose a fully flexible model which has two desirable featuresnot present in the usual econometric specifications implemented in applied research. First, itallows for flexible dynamics and for testing restrictions on these dynamics. Second, it does notimpose equivalence between alternative Parallel assumptions. We illustrate the usefulness of ourapproach by revising the results of several recent papers in which the difference-in-differencestechnique has been applied.The core assumption to identify the treatment effect in difference-in-differences estimators is theso-called Parallel Paths assumption, namely that the average change in outcome for the treated inthe absence of treatment equals the average change in outcome for the non-treated. We define afamily of alternative Parallel assumptions and show for a number of frequently used empiricalspecifications which parameters of the model identify the treatment effect under the alternativeParallel assumptions. We further propose a fully flexible model which has two desirable featuresnot present in the usual econometric specifications implemented in applied research. First, itallows for flexible dynamics and for testing restrictions on these dynamics. Second, it does notimpose equivalence between alternative Parallel assumptions. We illustrate the usefulness of ourapproach by revising the results of several recent papers in which the difference-in-differencestechnique has been applied
SN 2340-5031
YR 2012
FD 2012-12
LK https://hdl.handle.net/10016/16065
UL https://hdl.handle.net/10016/16065
LA eng
NO Bothauthors acknowledge financial help from the Spanish government through grant ECO2012-31358
DS e-Archivo
RD 23 jun. 2024