This preview shows page 1. Sign up to view the full content.
Unformatted text preview: age of classes skipped. o Smart students tend to have higher GPA other things equal • With panel data, one observe for the same student how GPA change when percentage of classes skipped change in different quarters. These changes in GPA cannot be caused by fixed student attributes. • Panel data lets us eliminate omitted variable bias when the omitted variables are individual/state/firm characteristics that are fixed over time. Panel Data with Two Time Periods 3 In the following we show how to use panel data to control for fixed effects and hence eliminate the omitted variable bias that would otherwise be caused by omitting those time invariant unobservables. Consider the panel data model, DrunkdriveRateit = β0 + β1BeerTaxit + β2wi + uit wi is a factor that does not change over time (culture, popularity of teenage drinking), at least during the years on which we have data. • Suppose wi is not observed, so its omission could result in omitted variable bias. • The effect of wi can be eliminated using T = 2 years. The key idea: Any change in the drunk driving rate from 1982 to 1988 cannot be caused by wi, because wi (by assumption) does no...
View Full Document
This document was uploaded on 03/11/2014.
- Spring '14