Hi everyone,
I came across some materials and they suggested that technically FE effect can be accounted for using dummies e.g.
xtreg y x1 x2, fe
is equivalent to
reg y x1 x2 i.country
when the panel used xtset country time.
Suppose x2 is a time invariant variable, if I include x2 into the FE model it will be omitted due to collinearity, however the coefficient can be estimated using the reg equation. So what is the implication for that coefficient? Is it an incorrect estimation?
Thanks in advance!
I came across some materials and they suggested that technically FE effect can be accounted for using dummies e.g.
xtreg y x1 x2, fe
is equivalent to
reg y x1 x2 i.country
when the panel used xtset country time.
Suppose x2 is a time invariant variable, if I include x2 into the FE model it will be omitted due to collinearity, however the coefficient can be estimated using the reg equation. So what is the implication for that coefficient? Is it an incorrect estimation?
Thanks in advance!
Comment