Working Paper: CEPR ID: DP10238
Authors: Li Dai; Patrick Minford; Peng Zhou
Abstract: We use available methods for testing macro models to evaluate a model of China over the period from Deng Xiaoping's reforms up until the crisis period. Bayesian ranking methods are heavily influenced by controversial priors on the degree of price/wage rigidity. When the overall models are tested by Likelihood or Indirect Inference methods, the New Keynesian model is rejected in favour of one with a fair-sized competitive product market sector. This model behaves quite a lot more 'flexibly' than the New Keynesian.
Keywords: Bayesian Inference; China; DSGE; Indirect Inference
JEL Codes: C11; C15; C18; E27
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Hybrid model (C59) | Better explanation of GDP dynamics (E20) |
| Hybrid model (C59) | Better explanation of inflation dynamics (E31) |
| Hybrid model (C59) | Better explanation of interest rates dynamics (E43) |
| Hybrid model (C59) | Greater flexibility (D29) |
| Greater flexibility (D29) | Better absorption of shocks (F35) |
| Lower response of inflation to shocks (E31) | More competitive product market (L19) |
| Hybrid model (C59) | Lower response of inflation to long-lasting shocks (E31) |
| Hybrid model (C59) | Reduced variance of price shocks (E39) |
| Hybrid model's lower Taylor rule responsiveness (C54) | Dampened investment responses (G31) |
| Hybrid model (C59) | More credible representation of macroeconomic behavior (E19) |