From a new working paper:
This paper provides statistical evidence suggesting that in industrial countries, recessions that are associated with either banking crises or housing crises dampen output far more than ordinary recessions… [W]e find that ordinary recessions are followed by strong recoveries that make up for almost all the preceding shortfall in output. This bounceback tends to be significantly smaller following recessions associated with banking crises or housing crises.
The authors also discuss other recent findings in the literature that have followed Reinhart and Rogoff in supporting this view:
Ignited partly by Reinhart’s and Rogoff’s (2008) attempt to draw lessons for the course of the US subprime crises from historical episodes with financial crises in other countries across the last centuries, a number of recent studies have investigated recessions triggered by severe crises and their aftermath…The general finding of these studies is that recessions triggered by severe crises are followed by rather weak and slow recoveries, and thus have long lasting effects on output.
It’s a good thing for Krugman he never took that bet with Mankiw.