IDENTIFICATION AND ESTIMATION OF TRIANGULAR SIMULTANEOUS EQUATIONS MODELS WITHOUT ADDITIVITY

被引：294
作者：
Imbens, Guido W. [1 ]
Newey, Whitney K. [2 ]
机构：
[1] Harvard Univ, Dept Econ, Littauer Ctr, Cambridge, MA 02138 USA
[2] MIT, Dept Econ, Cambridge, MA 02142 USA
来源：
ECONOMETRICA | 2009年 / 77卷 / 05期
关键词：
Nonseparable models; control variables; quantile effects; bounds; average derivative; policy effect; nonparametric estimation; demand analysis; INSTRUMENTAL VARIABLE ESTIMATION; NONPARAMETRIC-ESTIMATION; REGRESSION ESTIMATION;
D O I：
10.3982/ECTA7108
中图分类号：
F [经济];
学科分类号：
02 ;
摘要：
This paper uses control variables to identify and estimate models with nonseparable, multidimensional disturbances Triangular simultaneous equations models are considered, with instruments and disturbances that arc independent and it reduced form that is strictly monotonic in it scalar disturbance. Here it is shown that the conditional cumulative distribution function of the endogenous variable given the instruments is a control variable Also, for any control variable, identification results are given for quantile, average, and policy effects. Bounds are given when it common support assumption is riot satisfied. Estimators of identified objects and bounds are provided, and a demand analysis empirical example is given
引用
收藏
页码：1481 / 1512
页数：32
相关论文
共 53 条
[11]   An IV model of quantile treatment effects [J].
Chernozhukov, V ;
Hansen, C .
ECONOMETRICA, 2005, 73 (01) :245-261
[12]   Instrumental variable estimation of nonseparable models [J].
Chernozhukov, Victor ;
Imbens, Guido W. ;
Newey, Whitney K. .
JOURNAL OF ECONOMETRICS, 2007, 139 (01) :4-14
[13]   Nonparametric identification under discrete variation [J].
Chesher, A .
ECONOMETRICA, 2005, 73 (05) :1525-1550
[14]   Identification in nonseparable models [J].
Chesher, A .
ECONOMETRICA, 2003, 71 (05) :1405-1441
[15]  
CHESHER A, 2002, CWP2002
[16]  
Darolles S., 2003, NONPARAMETRIC INSTRU
[17]   Instrumental variables estimators of nonparametric models with discrete endogenous regressors [J].
Das, M .
JOURNAL OF ECONOMETRICS, 2005, 124 (02) :335-361
[18]  
DAS M, 2001, MONOTONE COMP STAT E
[19]   Identification of treatment effects using control functions in models with continuous, endogenous treatment and heterogeneous effects [J].
Florens, J. P. ;
Heckman, J. J. ;
Meghir, C. ;
Vytlacil, E. .
ECONOMETRICA, 2008, 76 (05) :1191-1206
[20]   Nonparametric methods for inference in the presence of instrumental variables [J].
Hall, P ;
Horowitz, JL .
ANNALS OF STATISTICS, 2005, 33 (06) :2904-2929
123456