Instrumental Variable Estimation of Nonlinear Errors-in-Variables Models
Instrumental Variable Estimation of Nonlinear Errors-in-Variables Models
This paper establishes that instruments enable the identification of nonparametric regression models in the presence of measurement error by providing a closed form solution for the regression function in terms of Fourier transforms of conditional expectations of observable variables. For parametrically specified regression functions, we propose a root n consistent …