Time-varying effects of extreme weather shocks on output growth of the United States

Abstract

This study investigates the impact of a structural shock to a metric of extreme weather, identified using sign restrictions, on output growth (and inflation) in the United States (US) from 1961 to 2022, using a new class of time-varying parameter vector autoregressive (endogenous TVP-VAR) model, whereby the identified structural innovation is allowed to influence the dynamics of the coefficients in the model unlike in traditional TVP-VARs. Our results provided evidence that severe weather shocks adversely affect output growth (and inflation) over the forecast horizon of one- to twelve-quarter-ahead. More importantly, we find that the effect of extreme weather on the US macroeconomic variables is indeed time-varying, with the impacts becoming smaller in recent times, possibly due to improved adaptation to climate change.