Section_6_Spring2008 - Ashenfelter and Krueger (1994) Econ...

Info iconThis preview shows pages 1–3. 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
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Ashenfelter and Krueger (1994) Econ 142 Spring 2008 Raymundo M. Campos-Vazquez February 27, 2008 Contents 1 Ashenfelter and Krueger (1994) 1 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Ignoring Biases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3.1 OLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3.2 GLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3.3 Measurement Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4 Solving Biases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4.1 Measurement Error: Average . . . . . . . . . . . . . . . . . . . . . . 4 1.4.2 Fixed E/ects - First Di/erencing . . . . . . . . . . . . . . . . . . . . 5 1.4.3 IV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.4.4 IV+Fixed E/ects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1 Ashenfelter and Krueger (1994) 1.1 Motivation Ashenfelter and Krueger (1994) is a famous and interesting paper in economics. The paper is famous because the methodology is clean and the identi&cation strategy is very interesting 1 (i.e. they collected their data, etc). I strongly recommend reading carefully this paper as well as to spend time understanding Problem Set 2. We are interested in the causal e/ect of education on earnings. Suppose we have the following regression: Y i = & + S i + e i (1) where Y is log wages and S is years of schooling. In order to &nd the causal e/ect of schooling on wages we need Cov ( e; S ) = 0 : This is a strong assumption. The assumption is not satis&ed because Schooling is not exogenous. You pick your own level of schooling, and you decide your level of schooling for some reason and probably this reason is correlated with other components that determine earnings at the same time. If this is the case, thewith other components that determine earnings at the same time....
View Full Document

Page1 / 6

Section_6_Spring2008 - Ashenfelter and Krueger (1994) Econ...

This preview shows document pages 1 - 3. Sign up to view the full document.

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