Identification in nonparametric simultaneous equations models

• 成果类型: Article
• 署名作者: Matzkin, Rosa L.
• 署名单位: University of California System; University of California Los Angeles
• 刊物名称: ECONOMETRICA
• ISSN/ISSBN: 0012-9682
• DOI: 10.3982/ECTA5940
• 发表日期: 2008
• 页码: 945-978
• 关键词: IDENTIFIABILITY CRITERIA; systems

摘要

This paper provides conditions for identification of functionals in nonparametric simultaneous equations models with nonadditive unobservable random terms. The conditions are derived from a characterization of observational equivalence between models. We show that, in the models considered, observational equivalence can be characterized by a restriction on the rank of a matrix. The use of the new results is exemplified by deriving previously known results about identification in parametric and nonparametric models as well as new results. A stylized method for analyzing identification, which is useful in some situations, is also presented.