Contrary to arguments advanced by others, it is shown in this note that, in general, a stochastic variation for a given relationship need not be accentuated in simultaneous models. Any statement to that effect must be based on a theory about shocks.