Working Paper: CEPR ID: DP15684
Authors: David Meenagh; Patrick Minford; Michael R. Wickens
Abstract: We ask whether Bayesian estimation creates a potential estimation bias as compared with standard estimation techniques based on the data, such as maximum likelihood or indirect estimation. We investigate this with a Monte Carlo experiment in which the true version of a New Keynesian model may either have high wage/price rigidity or be close to pure flexibility; we treat each in turn as the true model and create Bayesian estimates of it under priors from the true model and its false alternative. The Bayesian estimation of macro models may thus give very misleading results by placing too much weight on prior information compared to observed data; a better method may be Indirect estimation where the bias is found to be low.
Keywords: bayesian; maximum likelihood; indirect inference; estimation bias
JEL Codes: C11; E12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Maximum likelihood estimates (C51) | Biased estimates (C51) |
| Indirect estimation methods (C13) | More robust alternative to Bayesian estimation (C51) |
| Choice of prior distribution in Bayesian estimation (C11) | Posterior estimates of macroeconomic models (C51) |
| True model has high wage-price rigidity (C54) | Prior based on flexible wages and prices leads to biased posterior estimates (C51) |
| True model is based on flexible prices (C54) | Prior that assumes high rigidity leads to distorted estimates (C51) |
| Bayesian estimation process (C11) | Erroneous conclusions about underlying economic mechanisms (E19) |