“…Indeed, the excited state transition moments and energies for many states are often needed to accurately calculate the dispersion of the hyperpolarizability [33,34] and there are wellknown problems associated with extrapolating β 0 to zero frequency. [35,36] Even when only two electronic states dominate the response, the damping correction must be carefully taken into account [37,38] and vibronic overtones may complicate the analysis. [39,40] Secondly, there is nothing intrinsic about β 0 as we can argue from the fact that it can be large for very different underlying reasons.…”