An analysis using county-level US banking data
Ignazio Angeloni (HKS and SAFE)
Chantawit Tantasith (University of Oxford)
Johannes Kasinger (Goethe University Frankfurt)
The seminar presents preliminary results of a project analyzing whether bank behavior and performance help explain recent patters of poverty and inequality at the local level in the US.
The analysis links up with the recent literature on the deterioration of economic and social conditions in certain US regions. This literature has not considered financial factors so far. Yet, a long-standing stream of research has demonstrated that finance is an important co-determinant of economic performance and growth.
County-level data on banking conditions do not exist. However, county-level breakdown is necessary for this analysis, because there exist widely different economic and social conditions across different parts of the same states.
We developed a methodology to obtain, based on bank-level data from the FDIC, county-level estimates of a series of indicators of banking structure and performance (e.g., size, concentration, incidence of bank type (community, large BHC), capitalization, efficiency measures, interest margins, price of risk, NPLs, business model, sectoral specialization, etc.).
The outcome is a series of data panels, one for each year, each containing about 25 banking indicators for each of the 3000+ counties of the US. The data construction is now completed for 6 years (2000, 2005, 2010, 2018, 2019 and 2020).
The seminar presents three parts of our ongoing work:
- Our approach to constructing indicators of bank performance and banking structure at the level of individual countries.
- Descriptive evidence on the relationship between bank structures and performance on one side, and indicators of economic performance on the other, with particular emphasis on unemployment, poverty and inequality;
- Some preliminary panel estimates of this relationship.