Encouraged to Cheat? Federal Incentives, Career Concerns, and Local Freedom of Press as Determinants of Under-Reporting of COVID-19 Mortality at the Sub-National Level

This paper investigates the determinants and consequences of manipulating coronavirus statistics in an authoritarian federation; in particular, we look at how the career concerns of local politicians and the regional free press determine the under-reporting of COVID-19 mortality at the sub-national level in Russia. Our estimation strategy takes advantage of a natural experiment where the onset of the pandemic has adjourned the national referendum enabling new presidential terms for Vladimir Putin. To implement the rescheduled referendum, the Russian government needed the help of sub-national elites in manufacturing favorable COVID-19 statistics to convince the public that the pandemic is under control. While virtually all regions engaged in data manipulation, as we show by comparing the official data on COVID-19 mortality published in real-time and the excess mortality reported by the authorities only after the referendum, there was a substantial variation in the degree of misreporting. We show that a third of the variation can be explained solely by an exogenously set asynchronous schedule of governors’ elections, winning which depends almost exclusively on the support from the federal authorities. We find a robust causal relationship between the proximity to the governor’s election and the under-reporting of mortality associated with COVID-19 before the referendum, but not after when the political incentives from the federal center vanished. The local free press becomes a strong predictor of less under-reporting of COVID-19 mortality after the referendum. Looking at the consequences of the data manipulations, we find that under-reporting undermines individual trust in official statistics and decreases unresponsiveness of self-isolating behavior to changes in official COVID-19 mortality