Estimating the Impact of the Covid-19 Emergency on Tax Revenues in Guatemala: A Time Series Approach

Abstract

Applications of time series models serve two different purposes: (1) as forecasting techniques, they are used to project the trajectory of a variable of interest during a certain number of future periods; (2) in the analysis of interventions, they are used to evaluate the effect of a significant disturbance on the process being studied. We use both types of application to study monthly tax revenues in Guatemala. In Section 2 we use data for 2010-2019 in order to compare two alternative models: (a) the Box-Jenkins (ARIMA) model, and (b) the Holt-Winters exponential smoothing model. In Section 3 we use post-2019 data to estimate the fiscal effects of the emergency measures implemented to contain the Covid-19 pandemic.