Working Paper: NBER ID: w29281
Authors: Andrew Metrick; Paul Schmelzing
Abstract: We present a new database of banking-crisis interventions from the Roman Empire to the present, covering 1,946 interventions in 20 categories across 143 countries. We demonstrate that crisis-intervention patterns are significantly related to income and fiscal variables and to measures of the political system and currency regime. GDP losses following crises are economically significant and are larger for wealthier countries, with some evidence that these losses are mitigated by democratic political systems and liberal currency regimes. Finally, intervention frequencies reached an apex during the post-Bretton Woods era, continuing a secular increase since at least the late 17th century.
Keywords: No keywords provided
JEL Codes: G01; G28
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| government interventions (H53) | stability of the financial sector (E44) |
| government interventions (H53) | reduce likelihood of bank runs (E44) |
| reduce likelihood of bank runs (E44) | stabilize the financial system (G28) |
| increase in intervention sizes and frequencies (I24) | more aggressive approach to crisis management (H12) |
| more aggressive approach to crisis management (H12) | rising financial sector stress (E44) |