Impact of Covid-19 on Internet Usage: A Case Study of Pakistan

Abstract

This study is related to the investigation of impact of COVID-19 on internet usage. An ordinary least square method is implied on data collected from Jan-21 to Oct-21. The assume effect variable is total number of subscribers of broadband. Total death cases are assumed cause variable whereas newly reported cases and economic growth are moderating variables. Test for unit root shows all variables are found to be stationary at first difference. All variables are normally distributed evidence from Jarque-Bera probability. The Kurtosis value explains that the variables are playkutric. A positive correlation is observed between death cases and total internet subscribers. The regression model is significant and the variables explain 96%. The total deaths have significant positive impact on internet subscribers. The estimation result shows coefficient value of 0.74884. Granger causal test shows that new deaths cases do cause total number of subscribers. The model satisfy all necessary assumption of ordinary least square. There no auto correlation, multi collinearity, and heteroskedasticity in the model. Whereas the variables are not cointegrated with each other. The cusum and cusum of squares shows that the model is stable.