lect24_2010

lect24_2010 - 1 / 23 Introduction to Econometrics Econ 322...

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

View Full Document Right Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1 / 23 Introduction to Econometrics Econ 322 Fall, 2010 Lecture 24: Instrumental Variables Estimation (IV) I November 29, 2010 Topics Covered triangleright Topics Covered Motivation IV in the SLRM with one instrument Conditions for a valid instrument Two Stage Least Squares (2SLS) Derivation of 2SLS Inference with the 2SLS estimator General IV with multiple Instruments Identification Two Stage Least Squares in the GLRM A Digression: Simultaneous Equation Models Estimating the First Equation Estimating the Second Equation Estimating the Third Equation Example: Estimating a Demand Elasticity 2 / 23 1. IV 2. 2SLS 3. Identification Motivation Topics Covered triangleright Motivation IV in the SLRM with one instrument Conditions for a valid instrument Two Stage Least Squares (2SLS) Derivation of 2SLS Inference with the 2SLS estimator General IV with multiple Instruments Identification Two Stage Least Squares in the GLRM A Digression: Simultaneous Equation Models Estimating the First Equation Estimating the Second Equation Estimating the Third Equation Example: Estimating a Demand Elasticity 3 / 23 square So far we have discussed a number of threats to the “internal” validity of a regression model (GLRM). square These include: – omitted variable bias – simultaneous causality bias ( X causes Y and Y causes X) – errors in variables (X is measured with error) square We looked at a situation where we could exploit panel data to control for an omitted variable. We saw that if the omitted variable was either constant across the “cross-section” or constant across time then we can mitigate omitted variable bias by including “fixed” effects or “time” effects respectively. Motivation (Cont) Topics Covered triangleright Motivation IV in the SLRM with one instrument Conditions for a valid instrument Two Stage Least Squares (2SLS) Derivation of 2SLS Inference with the 2SLS estimator General IV with multiple Instruments Identification Two Stage Least Squares in the GLRM A Digression: Simultaneous Equation Models Estimating the First Equation Estimating the Second Equation Estimating the Third Equation Example: Estimating a Demand Elasticity 4 / 23 square If we are lucky and have panel data and we are lucky that the omitted variable does not vary across the “cross-section” or over time then we don’t get omitted variable bias. square However, sometimes we are not that lucky. If we – have an omitted variable that varies over time and across states or; – have an omitted variable and do not have panel data and cannot observe the omitted variable square Then we have omitted variable bias square If we have this case, or we have a problem of simultaneous causality bias (sometimes called regressor endogeneity ), or we have an errors in variable then we need to have a method that yields consistent estimates of the model’s parameters....
View Full Document

This document was uploaded on 10/26/2011 for the course ECON 327 at Rutgers.

Page1 / 23

lect24_2010 - 1 / 23 Introduction to Econometrics Econ 322...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online