Using WESML required an estimate of the proportion of companies in the population that manipulate earnings. Assuming that the popula-tion from which the companies were sampled is the population of Compustat companies, one estimate of the proportion of manipulators equals 0.0069 (50/7,231). Because I have no way of assessing the validity of this assumption, I also evaluated the sensitivity of the model to other specifications of the prior probability of manipulation.