Hi dear fellows,
I'm using fixed effect with panel data from 2011 to 2015 and my command is xtreg y x1 x2 i.year, fe. But it suffers serial correlation.One of the suggestions is that use First difference. Would the command be reg D.(Y x1 x2 year), nonconstant or reg D.(Y x1 x2 2011 2012 2013 2014 2015), nonconstant where 2011 indicates the dummy variable for year 2011.
In addition, how is the command reg D.(Y x1 x2 year) different from the command xtreg D(Y x1 x2), fd?
Thanks in advance.
Comment