The editorial of the new May 2022 issue of The Econometrics Journal celebrates ten years of Denis Sargan Econometrics Prizes. It recalls the wonderful sequence of past prize winners, in the hope and expectation that this will inspire the next generation of econometricians to send their best work to The Econometrics Journal.
The May issue's lead article studies causal mediation analysis in the presence of a (possibly) large number of observed potential confounders. It develops n-1/2-consistent and asymptotically normal estimators of the direct and indirect (via a mediator) treatment effects that are robust to some degree of misspecification and control for the confounders in a data-driven way. It does so by combining efficient-score based methods for robustly estimating direct and indirect treatment effects with Chernozhukov et al.'s (2018) double machine learning approach. It demonstrates the use of these estimators in an analysis of the effects of health insurance coverage on health in the NLSY, distinguishing between direct effects and effects mediated by routine checkups. The article was authored by Helmut Farbmacher, Martin Huber, Lukáš Lafférs, Henrika Langen, and Martin Spindler. Like all earlier Editors’ Choices of lead article, it is freely available from OUP.
The issue also contains a study from our recent Virtual Issue on COVID-19. This article by Thomas Dimpfl, Jantje Sönksen, Ingo Bechmann, and Joachim Grammig shows how official data on cases and fatalities can be combined with data from seroepidemiological studies in infection hotspots to estimate the infection fatality rate (IFR) of SARS-CoV-2. It then applies this method to estimate the IFR during Germany's first wave of COVID-19 and finds an IFR of 0.83%, which is way higher than the rates reported in earlier studies.
The May issue contains ten more fine articles, each illustrating its novel contributions to econometrics with an empirical application.