Comments on: Michael P. Keane 'Structural vs. atheoretic approaches to econometrics'

Richard Blundell, University College London, Department of Economics, London, critically comments on Michael P. Keane's article on 'Structural vs. atheoretic approaches to econometrics'. He analyzes that although a structural framework would seem to be a prerequisite for the interpretation of any policy reform, the degree to which a stochastic structure needs to be completely specified and empirically implemented will depend on the question to be answered. Keane recognizes experimental contrasts but the combination of experimental and observation evidence warrants further serious consideration as a creative way forward for building credible structural models. Keane suggests that structural discrete stochastic choice models considerations of non/semiparametric identification should be replaced by sample hold-out validation studies, to quote certain primacy for validation exercises over identification analysis for purposes of building confidence in models.