Impact of COVID-19 on Disaggregate Consumption and Online Retail Sales: Evidence from the USA

This study applies the difference-in-difference technique to analyze the consumption pattern during COVID-19 against pre-COVID-19 years. We analyze the online retail sales before and after COVID-19 using time series and linear regression models. Time series intervention analysis results suggest that COVID-19 has caused a statistically significant change in the mean level of online retail sales share in e-commerce. Using a difference-in-difference approach, we find a 4% decrease in aggregate consumption from March to December 2020 compared to the benchmark period although statistically insignificant. Further, using a fixed effects model with time dummies, we find a nearly 8% significant decrease in March–April and a 2% decrease in May–June, which is not significant maybe because the lockdown restrictions were lifted during that time. We infer that the aggregate consumption decreased during the strictest months of lockdown and COVID-19 had a heterogeneous impact across categories of consumption.