Hey statalist members,
I am trying to run two different regressions on panel data using the ivreg2 command. My first regression uses country-specific fixed effects by including dummy variables for the countries in the dataset. The problem occurs when I try to replace the dummy variables with some other factors. Literature suggests my regression could use some other variables as controls that might impact the dependent variable, instead of using the country dummies to account for country-specific fixed effects. These control variables are all time-invariant and take on only one variable for each country. I am getting the exact same results using both of these regressions. I have done research online on this and figured that since the variables are all time-invariant they can be seen as a multiple of the country dummy, meaning they both measure roughly the same thing. I wonder how I could test for the effects of the time-invariant variables on the dependent variable.
Any help would be appreciated!
Lukas Jannink
I am trying to run two different regressions on panel data using the ivreg2 command. My first regression uses country-specific fixed effects by including dummy variables for the countries in the dataset. The problem occurs when I try to replace the dummy variables with some other factors. Literature suggests my regression could use some other variables as controls that might impact the dependent variable, instead of using the country dummies to account for country-specific fixed effects. These control variables are all time-invariant and take on only one variable for each country. I am getting the exact same results using both of these regressions. I have done research online on this and figured that since the variables are all time-invariant they can be seen as a multiple of the country dummy, meaning they both measure roughly the same thing. I wonder how I could test for the effects of the time-invariant variables on the dependent variable.
Any help would be appreciated!
Lukas Jannink
Comment