ECON 151 Lecture Notes - Lecture 9: Econometric Model, Econometrics

• School: Santa Clara University
• Department: Economics
• Course: ECON 151
• Professor: Bouguen

Document Summary

Note that simple difference are estimated using the following model: Y = 0 + 1s + where 1 = dsc and estimated using ols. Sc estimator suffers from the same limitation than ols. Specifically, sc suffers from the zero conditional mean assumption (0cm) violation. Mathematically, ocm is violated when e( |s) = 0. 1 t = 1 is the performance of kids i at time t=1 (after schooling) 0 t = 1 s the performance of kids i at time t=0 (before schooling) Yi = 0 + 1 after i + where after takes value 1 when the data indicates after data. 1 = dba importantly, to estimate a before-after model, you usually need data to be in the long format. Wide vs. long format score t = b0 + b1*time + e. Both methods suffer from massive selection bias; selection bias = ocn. In reality, ba and sc are ols models;