Multidimensional Unobservables in Nonlinear Models

The talk will discuss several structural models with multidimensional unobservables.  New identification results for such models will be presented, together with estimation methods based upon constructive identification results. The methods allow counterfactual analysis for fixed values of the unobservable variables. Generalized regression models with multiple nonparametric indices, hedonic equilibrium models, and some interaction effects models will be dealt with in detail.