Bayesian Regression and the Expansion Method

This paper shows that within a Bayesian frame of reference the propensity to regard a model as invariant can be formalized into a prior probability density function of the parameters of an expanded formulation in which this model is encapsulated. The Bayesian approach allows to bring to the surface the implications of alternative levels of commitment to invariance assumptions on the results of empirical analyses, and can quantify the comparative strength of alternative drift specifications. The “expanded” Bayesian regressions are demonstrated by an example focusing upon the effectiveness of family planning.