Advanced Panel Data Methods

Chapter 14_Review

Course tutor: Ms. Le Thi Ngoc Mai 1

Fixed effects estimation
Fixed effect, potentially correlated
with explanatory variables

Form time-averages for

each individual

Because (the fixed effect is removed)

• Estimate time-demeaned equation by OLS (the unobserved time-

constant effect in is controlled)
Fixed effects estimation
• Example: Effect of training grants on firm scrap rate

Time-invariant reasons why one firm is more productive than another are controlled for.
The important point is that these may be correlated with the other explanatory variables.

Stars denote
Fixed-effects estimation using the years 1987, 1988, and 1989: time-demeaning

Training grants significantly improve productivity (with a time lag)

Random effects estimation
The individual effect is assumed to be “random” i.e.
completely unrelated to explanatory variables

Random effects assumption:

The composite error term = ai + uit is uncorrelated with the explanatory variables

• Transformed equation:

Quasi-demeaned data
Random effects estimation
• Example: Wage equation using panel data

Random effects is used because

many of the variables are time-
invariant.
Random effects or fixed effects?
• Fixed effects allows arbitrary correlation between and the
explanatory variables, while random effects does not => FE is widely
thought to be a more convincing tool for estimating ceteris paribus
effects.
• Still, random effects is applied in certain situations. Most obviously, if
the key explanatory variable is constant over time, we cannot use FE
to estimate its effect on y.
• It is still fairly common to see researchers apply both random effects
and fixed effects, and then formally test for statistically significant
differences in the coefficients on the time-varying explanatory
variables => Hausman test