Working Paper: NBER ID: w15379
Authors: Stephen G. Cecchetti; Marion Kohler; Christian Upper
Abstract: We study the output costs of 40 systemic banking crises since 1980. Most, but not all, crises in our sample coincide with a sharp contraction in output from which it took several years to recover. Our main findings are as follows. First, the current financial crisis is unlike any others in terms of a wide range of economic factors. Second, the output losses of past banking crises were higher when they were accompanied by a currency crisis or when growth was low at the onset of the crisis. When accompanied by a sovereign debt default, a systemic banking crisis was less costly. And, third, there is a tendency for systemic banking crises to have lasting negative output effects.
Keywords: Financial Crises; Economic Activity; Output Losses
JEL Codes: E32; E44; G01
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| systemic banking crises (F65) | significant output losses (E23) |
| systemic banking crises + currency crisis (F65) | longer and deeper contractions (J59) |
| initial growth is low (O41) | longer and deeper contractions (J59) |
| sovereign debt defaults (H63) | mitigate costs of banking crisis (F65) |
| lower growth preceding a crisis (F44) | longer contractions (J59) |
| lower growth preceding a crisis (F44) | lower trough in output (E23) |
| systemic banking crises (F65) | lasting negative effects on GDP levels (F69) |