The research examines how limited formal education among females influences both the incidence and underreporting of rape cases in South Africa. The study uses a panel dataset comprising 244 municipalities observed annually between 2008 and 2024. Unlike previous studies, we assume that the observed number of reported rape cases is an outcome bounded by a stochastic reporting frontier. Deviations from this frontier are interpreted as underreporting. The analysis is performed using an instrumental variable stochastic frontier model and the maximum likelihood estimator. We find the average ratio of actual to potential rape cases approximately 0.66 implying that 34 percent or roughly one-third of rape cases goes unreported in South Africa. Approximately 49 percent of the observed variation in official rape cases is due to underreporting. Rape cases and underreporting are higher in municipalities where majority of females have no formal education. These findings are consistent with the proposition that females with limited formal education are more likely to normalize coercive sexual behaviour and conceal rape. The average underreporting of 34 percent additionally suggests that the social burden of rape in South Africa may be higher than official statistics suggest.
Brian Tavonga Mazorodze (Sat,) studied this question.