Temporal trends and seasonal patterns of violent crime in France

Violent behavior is a major public health concern, yet whether different forms of violent crime follow similar or distinct seasonal patterns remains unclear. We analyzed monthly national counts of police-recorded violent offenses in France from January 2016 to December 2024 using open-access data from the Ministerial Statistical Service for Internal Security. Offenses included homicides and attempted homicides, physical violence, sexual violence, robbery with a weapon, and violent robbery without a weapon. Seasonal patterns were assessed using quasi-Poisson generalized additive models with monthly modeled as a cyclic spline and adjustment for long-term trends. Sensitivity analyzes examined alternative April/COVID-19 specifications and the influence of the July 2016 Nice terrorist attack. Exploratory harmonic analyzes assessed a possible bimodal pattern. Overall, 4,756,265 violent offenses were recorded, including 9,161 homicides and attempted homicides. Physical violence increased by 62.4%, sexual violence by 136.9%, and homicides and attempted homicides by 28%, whereas violent robberies declined by 46.6%. In the original homicide model, seasonality was significant (p = 0.039) but was no longer significant after correction for the Nice attack (p = 0.545) or exclusion of 2016 (p = 0.399). No significant seasonality was observed for physical or sexual violence across sensitivity analyzes, while robbery results were model-sensitive. Exploratory analyzes suggested peaks in early summer and December–January, with borderline permutation evidence. Violent crime increased substantially in France, mainly through rises in physical and sexual violence. The apparent summer seasonality of homicide was not robust to sensitivity analyzes, and the possible bimodal pattern requires confirmation in longer time series. • Police-recorded violent crime in France increased substantially from 2016 to 2024, driven mainly by rises in physical and sexual violence. • The initially observed summer seasonality of homicides and attempted homicides was strongly influenced by the July 2016 Nice terrorist attack and was not robust in sensitivity analyzes. • Exploratory analyzes suggested a possible bimodal pattern, with increases in early summer and around December–January, but this finding requires confirmation. • Seasonal patterns differed across offense categories: no robust seasonality was observed for physical or sexual violence, while robbery findings were sensitive to model specification.