Solvatochromic shifts in molecular spectra are often used to determine excited-state dipole moments. In comparing results with the mon accurate gas-phase Stark effects studies, large discrepancies are obtained in the few cases in which both types of studies have been carried out. We attempt to reconcile the two techniques by reanalyzing solvatochromic shift measurements in the situation in which a field-induced perturbation by a nearby state is possible. It is found that, in the high field limit, if we reinterpret solvatochromic shift studies as measuring an effective dipole moment which is a function of the dipole moments of both states and the transition moment between the two states, a possible explanation of the discrepancy is obtained. This is demonstrated to be reasonable in the only case (aniline) for which sufficient data are available. We also include examination of field-induced intensity borrowing as well as certain cavity-induced resonances and magnetic field induced effects. Finally we suggest some experiments to test the extent to which this reinterpretation is needed.