I am using casscf method at gaussian 09 and I set pop=full to see the energies of the molecular orbitals. In the output file the calculation gives me only the kinetic energies (alpha) while for Orbital energies gives me 0.0000.
This question may be a bit old, but here's an answer, anyway, if anyone comes across:
the orbital energies you get from single-reference methods are eigenvalues of the Fock or Kohn-Sham equations. In multireference methods, that ceases to be the case, therefore the resulting orbitals do no longer have the property "energy" since they are in general not eigenfunctions of an operator anymore.