AI agents are transforming economic research. A new NBER working paper introduces DRIL, a methodology using artificial intelligence to automatically construct economic datasets from public sources. Researchers applied the system to build a Global Tax Expenditures Database for eight Latin American and Caribbean countries, generating 129 sources and 136 evidence records at minimal cost. The breakthrough demonstrates how automation can dramatically reduce expenses associated with dataset construction, traditionally one of empirical economics' most resource-intensive tasks. This advancement could fundamentally reshape how economists conduct research by lowering barriers to data collection and enabling faster, more comprehensive analysis across regions and time periods.
Post from MarketNews_en
Log in to interact with content.