Estimating the effect of maternal stress on birth outcomes in Chile: a pseudo-panel approach
Ein Vortrag von Serafima Chirkova (University of Chile) im Rahmen der Seminarreihe des AB Ökonomie am IOS.
Datum: 13. Februar 2018
Zeit: Beginn 10 Uhr!
Ort: Leibniz-Institut für Ost-und Südosteuropaforschung (IOS), Landshuter Str. 4 (Raum 109)
We study how prenatal maternal stress, caused by sustained seismic activity, affects birth outcomes in Chile. Using administrative data on the Chilean births during the period 2005-2015, we exploit spatial and temporal variation in exposure to earthquakes across Chile. First, assuming the random occurrence of earthquakes we apply the standard estimation strategy comparing affected and not affected groups. The problem of this approach is that it does not account for an unobserved heterogeneity (genetic predisposition, health or environmental conditions, etc.). If so, a pregnant women may not be randomly allocated between affected and not affected groups, as existing literature typically assumes, and it may result in biased estimators. To explore this possibility, we use the interactive-effects estimator, proposed by Bai (2009), on the constructed cohort pseudo-panel. Our approach allows us to consider the possibility that common shocks can have heterogeneous effects on each cross-sectional unit via the individual specific unobservable factors. Hence, the interactive-effects estimator provides more accurate estimates of the impacts that maternal characteristics have on birth outcomes. Our empirical evidence suggests that exposure to sustained seismic activity in early pregnancy increases (on average) newborns’ birth weight and has no effect on gestational length.