Supply-Side Effects and Role of Sentiment during COVID-19: Evidence from China

By leveraging a difference-in-difference (DID) framework, I illustrate that COVID-19 has led to a sustained and pronounced disruption of various supply-side variables across all 31 provinces in China. Particularly, con- trolling for non-pharmaceutical public health interventions (NPIs), more severely affected provinces would have substantially more loss-making firms, less production and inventories, and devaluation of assets. While the effect on profit, production, and inventory was gradually weakened, the pandemic led to a more persistent erosion of the firms' balance sheets. I then look at the role of social media during the pandemic by analyzing the sentiments of geocoded tweets of Sina Weibo (i.e., China's Twitter) during February 2020, the most distressing period in China. Aggregating sentiment labels to the province level, I found that regions with more positive sentiments had considerably newer establishment of industrial enterprises. The results are robust across various specifications of controls, especially local pandemic severity and the intensity of NPIs.