Authoritarian regimes' propensity to manipulate Covid-19 data: a statistical analysis using Benford's Law

There have been numerous claims that some countries do not reveal their reported cases of COVID-19 truthfully. Some studies suggest that autocratic countries tend to manipulate their reported figures. To test these claims, the article uses Benford's law (BL) to detect anomalies in the reported numbers of COVID-19. Many studies have used BL to test the reliability of the data recorded by countries and conclude that the data follows BL in most cases. This study applies BL to a set of 34 countries using Pearson's X-2 , Kuiper and mean absolute deviation tests to analyse the daily reported cases; then, correlating the tests results with four freedom indices using ordinary least square. The aim of the article is to demonstrate how countries with low freedom scores are likely to manipulate their COVID-19 reporting, whereas countries with high freedom scores do not show a manipulation in their data. The study finds that all indices show strong correlation to the degree of data manipulation, and with the World Press Freedom index showing the strongest relationship.