Abstract:
Bayesian econometrics provides a tidy theory and practical methods of comparing and combining several alternative, completely specified models for a common data set. It is always possible that none of the specified models describe important aspects of the data well. The investigation of this possibility, a process known as model validation or model specification checking, is an important part of applied econometric work. Bayesian theory and practice for model validation are less well developed. A well-established Bayesian literature argues that non-Bayesian methods are essential in model validation. This line of though persists in Bayesian econometrics as well; the paper reviews these methods. The paper proposes an alternative, fully Bayesian method of model validation based on the concept of incomplete models, and argues that this method is also strategically advantageous in applied Bayesian econometrics.
