A detailed decomposition for nonlinear econometric models

摘要: This paper proposes a detailed decomposition for nonlinear econometric models, covering well-known limited dependent variable models such as the logit, probit and Tobit models. The purpose of proposed method is to relate between-group differences in an outcome observable characteristics. overcomes disadvantages existing decompositions discussed literature. In particular, it leads unique absorbs present nonlinearities natural way. this paper, theoretically identified shown how can be estimated inference conducted. It then compared methods. An empirical application presented where results from are with demonstrated that linear methods Oaxaca-Blinder may not appropriate presence substantial nonlinearities, approach sensible differential.