Time and country dummies for cross-sectional regression
Posted: Fri Oct 05, 2012 12:15 pm
Hi there,
I am quite unfamilar with Eviews and so I have a couple of questions:
I have the following problem:
I have cross sectional data, where the dependent variable is e.g. abnomal (i.e. above industry average) ebitda margin. However, my main explanatory variable is supposed to be dummy variable taking the value of 1 or 0 (the series name is F1) depending on the characteristics of the company. As far as I know I simply inlcude @expand(f1, @dropfirst) in order to avoid the dummy variable trap. However, since the Ebitda margin is from different years (2002-2008) as well as countries (United Kingdom, France, Germany). So I am wondering now whether I should create a dummy series for each year seperately and then simply exclude one of them in oder to avoid the trap or whether i can also generate a dummy series for all at one (which as far as I know would not be a dummy anymore, since a dummy only takes values 1/0).
If I go for the first option (seperate dummies for each year) would a command like this to estimate the regression be correct:
log_egr c f1 log_em0 log_as0 log_gdp spa first y2 y3 y4 y5 y6 y7 y8 y9 z_uk z_fr
where as f1 corresponds to my main explantory variable and y2 y3 , z_uk, z_fr to the respective dummy variables.
Moreoever, i was wondering whether the order of explanatory variables is important, I mean that dummy variables have be after the other explanatory variables in the estimation equation.
Sorry for so many questions at once, I am reallly confused at the moment!
Thank you very much in advance for your help!
Regards,
Sili
I am quite unfamilar with Eviews and so I have a couple of questions:
I have the following problem:
I have cross sectional data, where the dependent variable is e.g. abnomal (i.e. above industry average) ebitda margin. However, my main explanatory variable is supposed to be dummy variable taking the value of 1 or 0 (the series name is F1) depending on the characteristics of the company. As far as I know I simply inlcude @expand(f1, @dropfirst) in order to avoid the dummy variable trap. However, since the Ebitda margin is from different years (2002-2008) as well as countries (United Kingdom, France, Germany). So I am wondering now whether I should create a dummy series for each year seperately and then simply exclude one of them in oder to avoid the trap or whether i can also generate a dummy series for all at one (which as far as I know would not be a dummy anymore, since a dummy only takes values 1/0).
If I go for the first option (seperate dummies for each year) would a command like this to estimate the regression be correct:
log_egr c f1 log_em0 log_as0 log_gdp spa first y2 y3 y4 y5 y6 y7 y8 y9 z_uk z_fr
where as f1 corresponds to my main explantory variable and y2 y3 , z_uk, z_fr to the respective dummy variables.
Moreoever, i was wondering whether the order of explanatory variables is important, I mean that dummy variables have be after the other explanatory variables in the estimation equation.
Sorry for so many questions at once, I am reallly confused at the moment!
Thank you very much in advance for your help!
Regards,
Sili