Using Integrated Non-linear ? Method To Estimate The Elasticity Of Intertemporal Substitution

The linear regression model has played an important role in economic analysis.The model has been widely used for its intuitiveness and explanatory power.The perfection of the theory and method system of linear regression makes it possible to obtain linear conditions through some transformation methods when dealing with some nonlinear problems.When suffered from endogeneity,the mature solution is the IV method.The IV estimate can easily use the exogeneity to obtain the moment condition.Therefore,the model can be estimated by using the GMM method.The core of the application of the GMM method is a series of moment conditions implied by the model.Under this framework,full use of the information contained in all moment conditions is the key to a consistent estimation.In empirical applications,the implicit condition of the model is often the conditional moment constraint.For such model estimation,especially when the condition contains instrumental variables,one method of estimating is to perform the direct reference to the unconditional moment constraint,if it is traditional.TSLS does not use the complete information contained in the instrumental variables;the other is based on the eigenfunction idea to use the set of IVs in the IV space to form an equivalent continuous moment condition for GMM estimation.The integral nonlinear IV method proposed in this paper uses a non-linear transformation function,so that it can capture more information that is not limited to the linear correlation information,and uses a unique weighting matrix,making the most of IV information.The degree of utilization of information is improved,thereby achieving a good estimation accuracy of the INI estimator.