Working Paper: CEPR ID: DP16054
Authors: Eric Monnet; Angelo Riva; Stefano Ungaro
Abstract: We investigate the causal impact of bank runs by exploiting a key feature of the French Great Depression (1930-1931) that created exogenous geographical variations in the withdrawals of bank deposits. Unregulated commercial banks coexisted with government-backed saving institutions (Caisses d’épargne). During the crisis, depositors who had an account in Caisses d’épargne were more likely to withdraw from banks. Pre-crisis density of Caisses d’épargne accounts was unrelated to economic and bank characteristics. Using this variable as an instrument, we find that a 1% decrease in bank branches reduced aggregate income by 1%. Our identification highlights how a shift of deposits towards safer institutions can affect financial fragility. It holds lessons for current financial regulation andthe design of central bank digital currency (CBDC).
Keywords: bank runs; flight-to-safety; banking panics; great depression
JEL Codes: E44; E51; G01; G21; N14; N24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| decrease in the number of bank branches (G21) | decrease in local GDP growth (F69) |
| pre-crisis density of caisses d'épargne accounts per capita (N13) | likelihood of bank runs (E44) |
| bank runs (E44) | decline in banking activity (G21) |
| bank runs (E44) | drop in real GDP (E20) |