%0 Work %T Essays on Identification in Econometric Models %A Tatiana Komarova %D 2008-10-24 %8 2018-09-16 %R http://localhost/files/ws859f784 %X This dissertation consists of three essays on the identification analysis of econometric models. The first essay explores the identification question in semiparametric binary response models when all regressors have discrete support. I suggest a recursive procedure that finds sharp bounds on the parameter of interest and can be applied to the analysis of identification sets in other single-index models. Furthermore, I investigate asymptotic properties of estimators of the identification set. I also propose three approaches to address the problem of empty identification sets when a model is misspecified. Finally, I present a Monte Carlo experiment and an empirical illustration to compare several estimation techniques. The second essay proposes an approach to proving nonparametric identification for the distributions of bidders' values in asymmetric second-price auctions. I consider the case where bidders have independent private values, and the only available data pertain to the winner's identity and to the transaction price. I provide conditions on observable data sufficient to guarantee point identification. I demonstrate how the techniques of the identification proof can be used to obtain identification in generalized competing risks models. Finally, I provide a sieve minimum distance estimator that consistently estimates the underlying valuation distribution of interest. The third essay analyzes identification in second-price auction within the private values framework when bidders' values are not independent. When only the winner's identity and the winning price are observed, neither the joint nor the marginal distribution of bidders' values are point identified but I derive tight bounds on the distribution of values for any subset of bidders. In addition, I investigate how these bounds change when data on other elements of the model become available. Finally, I use the representation of joint distributions through copulas and prove point identification for certain classes of copulas. %[ 2018-09-16 %9 Dissertation %~ Arch : Northwestern University Institutional Repository %W Northwestern