The Kubo formula for Hall conductance in the Quantum Hall Effect is given as $$\sigma_{xy}=i \hbar \sum_{n \neq 0}\frac{\langle{0}|J_y|n\rangle \langle n|J_x|0\rangle - \langle{0}|J_x|n\rangle \langle n|J_y|0\rangle}{(E_n-E_0)^2}$$ I have read (and understood) the derivation. However, I am still looking for an intuitive explanation for the formula, i.e. I am searching for answers to questions like:
1) Why do the energy eigenfunctions which are further away from $E_0$ contribute less to the sum?
2) What exactly is the meaning of 'crossing' behavior of the nominator?
3) And why does the $n=0$ state not contribute to the sum (apart of course from the fact that we had a divergence in the denominator and the nominator is $=0$)? Normally, only the unfilled bands can carry a current...