Page 1 of 1

Random effects estimation with an explicit formulation of equation

Posted: Wed Jul 10, 2019 12:11 am
by Michael_MA
Dear Colleagues,

I am currently trying to estimate a set of panel equations with explicit restrictions on parameter values in a following form:

log(Y) = c(1)*log(X) + c(1)*log(Z) + (1-c(1))*log(X0) + (c(1)-1)*c(2)*time + c(3).

However, when I try to include random cross-section effects, EViews generates the following message:

Random effects may not be estimated in specifications
involving an explicit equation. You must specify your
equation as a dependent variable followed by a list of
regressors.


By the way, such a problem is not observable in the case of fixed effects.

Does this imply that the random effects estimation with explicit parameter restrictions is not possible in EViews?