The Role of Delayed Data in the COVID-19 Pandemic

Abstract

Since the COVID pandemic began, data has been at the heart of debates over how to respond. However, it is not yet well understood how the lack of high-quality data affects policy decisions. In this paper, we focus on one specific aspect of the errors in COVID-19 data: delays in data reporting. We present one of the first lines of evidence for significant reporting delays in state-level COVID-19 data, and for heterogeneity in reporting delays across states and over time. We then use policy analysis of state-level interventions to illustrate how failing to consider reporting delays in data analysis can severely distort policymakers' decisions. Our analysis traces out heterogeneity in state reporting due to reliance on outdated fax-based reporting methods, and infers the severity of delay from the use of these methods and substantial short-term fluctuations in reported numbers.

Our results show that accounting for potential reporting delays in statistical analysis leads to sharp changes in the estimated effects of almost all policies we investigate. It is critical to take potential reporting delays into consideration when analyzing reported COVID-19 data, and there are large gains from systematically addressing the root causes of reporting delays in IT systems in general.

Note: Funding Statement: MIT Sloan Health Systems Initiative

Declaration of Interests: Both authors do not have conflicting interests in this paper.

Keywords: COVID-19, reporting delay, data quality

JEL Classification: M15