Identification and Estimation of a Nonseparable Production Function

Abstract

COVID-19 leads to a substantial decrease in the Chinese economy in early 2020. More importantly, the economy's recovery is not immediate; even new COVID-19 cases drop (close) to zero for a few months. I develop a nonseparable production function model to show productivity channel is a resolution to this puzzle. I first establish nonparametric identification and estimation of all the unspecified elements and provide estimators' asymptotic properties. Applying the model to the data, I find COVID-19 lowers current period productivity by 0.101, which then decreases current period output by 2.6%, causing the substantial economic decrease. The decrease of current period productivity reduces next period productivity by 0.084. The low current and next period productivity decrease next period output by 1.9%, causing the slow economic recovery. My estimation results also show that the more the new COVID-19 cases are, the larger the decrease on current and next period productivity and output.