What symmetries must be broken in a material in order for a certain optical effect, known as Kerr rotation, to appear? Kerr rotation has been observed in some high temperature superconductors, but it was uncertain as to how to interpret these measurements and identify the material symmetries responsible for the effect. Elaborating on previous discussions (cf. e.g. PRB 90, 035135 (2014)) the author offers a proof based on the reciprocity theorem that the polar Kerr effect can be only observed if the system breaks time-reversal symmetry.