lecture_note_112916

lecture_note_112916 - Econ 5280 Applied Econometrics...

Info icon This preview shows pages 1–5. Sign up to view the full content.

View Full Document Right Arrow Icon
Econ 5280: Applied Econometrics Instructor: Professor Jin Seo Cho ©Jin Seo Cho, 2016, All Rights Reserved.
Image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Chapter 5 Linear Models with Endogenous Errors 5.1 Introduction Note that the previous sections focus on the estimation of conditional expectation: E ( Y t X t ) , and the errors generated by the conditional expectation is orthogonal to the explanatory variable by its nature: E ( U t X t ) = 0 . This condition often does not hold for economic equations. In many cases, it is more appropriate eco- nomic equation that is not necessarily a conditional expectation. For example, demand function or supply function is not constructed by conditional expectation. It is an equation derived from economics that could be irrelevant to conditional expectation. For such a case, it is more appropriate to assume that errors are not orthogonal to the explanatory variables: E [ U t X t ] 0 . If this is the case, the OLS estimator cannot be a proper estimator, and the equation is often called the structural equation . The OLS estimator cannot consistently estimate the parameters in the structural equation. We may suppose that the following structural equation: Y t = X t β * + U t such that E [ U t X t ] 0 , and we may estimate the unknown parameter β * by the OLS estimation. Then, ̂ β n = β * + ( n - 1 n t = 1 X t X t ) - 1 ( n - 1 n t = 1 X t U t ) , where n - 1 n t = 1 X t X t a.s. E ( X t X t ) and n - 1 n t = 1 X t U t a.s. E ( X t U t ) 0 . 60
Image of page 2
Thus, it follows that ̂ β n a.s. β * + E ( X t X t ) - 1 E ( U t X t ) β * . This implies that the OLS estimator is not consistent for β * . We often call this that the OLS estimator is asymptotically biased . Due to this asymptotic bias, the OLS estimator has to be replaced by another estimator, and we consider the replacement in this chapter. 5.2 Instrumental Variable Estimator The instrumental variable (IV) estimator can be a proper estimator for β * . Before examining the precise implication of the IV estimator, we first provide the following assumptions for the IV estimation and its inference. 1. Dateset is a sequence of IID observations: {( U t , X t , Z t ) R 1 + 2 k t = 1 , 2 , ,n } ; 2. For some unknown β * , Y t = X t β * + U t such that E ( U t Z t ) = 0 ; 3. E ( U 4 t ) < , E [ X 2 t,i ] < , and E [ Z 4 t,i ] < for i = 1 , 2 , ,k ; 4. E [ Z t X t ] and n t = 1 Z t X t are positive definite; and 5. Σ = E [ U 2 t Z t Z t ] is positive definite. 5.2.1 IV Estimator and Its Asymptotic Properties The IV estimator is defined by the orthogonality condition. Note that the orthogonality condition E ( U t Z t ) = 0 ) can be manipulated as follows: 0 = E ( Z t U t ) = E [ Z t ( Y t - X t β * )] , so that β * = E ( Z t X t ) - 1 E ( Z t Y t ) . Here, E ( Z t X t ) - 1 exists because E ( Z t X t ) is positive definite. Furthermore, E ( Z t Y t ) also exists mainly because E ( Z t X t ) exists and Y t = X t β * + U t . Thus, we may estimate β * by the sample analogs in the right 61
Image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
side of this equation: ̃ β n = n - 1 n t = 1 Z t X t - 1 n - 1 n t = 1 Z t Y t = ( Z X ) - 1 Z Y , where Z = [ Z 1 , , Z n ] .
Image of page 4
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern