This preview shows pages 1–6. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: ECON 103, Lecture 15B: Instrumental Variables II Maria Casanova May 28th (version 0) Maria Casanova Lecture 15B Requirements for this lecture: Chapter 12 of Stock and Watson Maria Casanova Lecture 15B 1. IV regression with 1 regressor and 1 instrument We start from the population regression: Y i = + 1 X i + i , where X i and i are correlated. IV regression uses an instrumental variable Z to isolate the variation in X that is not correlated with . For an instrumental variable Z to be valid, it must satisfy two conditions: 1 Instrumental relevance: variation in the instrument is related to variation in X i , Corr ( Z i , X i ) 6 = 0 2 Instrumental exogeneity: the part of variation in X i captured by the instrumental variable is exogenous, i.e. not correlated with the error term. Corr ( Z i , i ) = 0 Maria Casanova Lecture 15B 1. IV regression with 1 regressor and 1 instrument If Z satisfies the two conditions to be a valid instrument, then 1 can be estimated using an IV estimator called twostage least squares (TSLS). As its name suggests, TSLS proceeds in two stages: 1 First, we isolate the part of X that is uncorrelated with by regressing X on Z using OLS: X i = + 1 Z i + v i Because Z i is uncorrelated with i , also + 1 Z i is uncorrelated with i . We dont know and 1 , so we estimate them and then compute the predicted values of X i , i.e. X i = + 1 Z i Maria Casanova Lecture 15B 1. IV regression with 1 regressor and 1 instrument 2 Second, we replace X i with X i in the regression of interest, and regress Y i on X i using OLS: Y i = + 1 X i + i Because X i is uncorrelated with i , the first least squares assumption holds. Thus the estimate 1 obtained by OLS in the second regression is consistent....
View
Full
Document
 Spring '07
 SandraBlack

Click to edit the document details