COVID-19 Uncertainty Index in Japan: Newspaper-Based Measures and Economic Activities

Measuring uncertainty and its economic impact are of major concern during the unprecedented crisis triggered by the coronavirus disease 2019 (COVID-19) pandemic. This paper constructs a newspaper-based measure that captures the uncertainty induced by COVID-19 and examines its economic impacts using a structural VAR model applied to Japanese data. We develop two types of uncertainty indices and identify two types of structural shocks in the VAR model: one measuring an epidemiological uncertainty, the other a policy-related uncertainty. Our findings are summarized as follows. First, the constructed series of uncertainty shows a spike after COVID-19 related events, indicating that our indices work well as a measure of COVID-19 induced uncertainty. Second, stock market variables show statistically significant responses to a policy-related uncertainty shock rather than an epidemiological uncertainty shock. Third, in contrast, real variables such as mobility and consumption tend to respond significantly to an epidemiological uncertainty shock. These findings highlight the importance of considering different types of uncertainty in order to properly assess the impact of COVID-19 induced uncertainty on economic activity.