I use a stacked first differences model to estimate the impact of globalization on unemployment in Western Europe.
• Panel data: 16 countries, yearly observations for the years 1995-2007
• 2SLS regression
• The following code shows the first-stage regression, where...
- I control for the same variables (c1, c2, c3, c4) and
- use time and country fixed effects
...just like in the reduced form / second-stage regression
Code:
xi: xtreg x z c1 c2 c3 c4 i.year, fe vce(cluster country)
- estimate for the instrument z: 1.6788Hence, the coefficient is highly significant.
- standard error of the instrument z: 0.005
To get the F-statistic, I use the following code:
Code:
test z
If not, what could have gone wrong?
I appreciate your help!
0 Response to First Stage Regression - Huge F-statistic
Post a Comment