Working Paper: CEPR ID: DP7508
Authors: Xiaoming Cai; Wouter Den Haan
Abstract: Several papers that make forecasts about the long-term impact of the current financial crisis rely on models in which there is only one type of financial crisis. These models tend to predict that the current crisis will have long lasting negative effects on economic growth. This paper points out the deficiency in this approach by analyzing the ability of "one-type-shock" models to correctly forecast the recovery from past economic downturns. It is shown that these models often overestimate the long-run impact of recessions and that slightly richer models that allow the effects of recessions to be both persistent and transitory predict recoveries much better.
Keywords: Financial Crisis; Forecasting; Great Recession; Unit Root
JEL Codes: C51; C53; E37
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| one-type-shock models (C69) | overestimate long-term negative consequences of recessions (E71) |
| one-type-shock models (C69) | pessimistic forecast of GDP recovery (E27) |
| components model (D10) | faster convergence towards new long-term growth path after downturn (F62) |
| one-type-shock models (C69) | biased forecasts of long-term growth (F17) |
| incorporating various economic indicators (E01) | enhance forecasting accuracy (C53) |