Understanding Bank Runs Through Historical Newspapers
Large language models are used to extract bank run data from historical newspapers, creating a comprehensive database for understanding U.S. banking crises.
Sergio Correia, Stephan Luck, and Emil Verner utilized large language models to extract information on bank runs from digitized historical newspaper articles. This effort has resulted in the most comprehensive database of bank runs in U.S. history, allowing users to explore individual episodes and original newspaper reports.