Keeping a tally of electoral fraud is difficult because of the very characteristics of the process and of the conditions surrounding how it is publicized, but also because of the complexity of a quantitative analysis of electoral anomalies. Quantitative approaches to these practices are rare and their results fragile. However, many political actors put forward estimates of the level and of evolutions of electoral fraud. The main challenge is to diagnose the current state of democracy, which is supposed healthier if fraud emerges less frequently. The overall aim is to enforce political strategies of legitimization or disqualification. Ultimately, these valuations yield less information on the phenomenon itself than on the actors who claim to measure it.