How do Google searches for symptoms, news and unemployment interact during COVID-19? A Lotka-Volterra analysis of google trends data.

In this paper I exploit Google searches for the topics "symptoms", "unemployment" and "news" as a proxy for how much attention people pay to the health and economic situation and the amount of news they consume, respectively. I then use an integrable nonautonomous Lotka-Volterra model to study the interactions among these searches in three U.S. States (Mississippi, Nevada and Utah), the District of Columbia and in the U.S. as a whole. I find that the results are very similar in all areas analyzed, and for different specifications of the model. Prior to the pandemic outbreak, the interactions among health searches, unemployment searches and news consumption are very weak, i.e. an increase in searches for one of these topics does not affect the amount of searches for the others. However, from around the beginning of the pandemic these interactions intensify greatly, suggesting that the pandemic has created a tight link between the health and economic situation and the amount of news people consume. I observe that from March 2020 unemployment predates searches for news and for symptoms. Consequently, whenever searches for unemployment increase, all the other searches decrease. Conversely, when searches for any of the other topics considered increase, searches for unemployment also increase. This underscores the importance of mitigating the impact of COVID-19 on unemployment to avoid that this issue swallows all others in the mind of the people.