In recent years, a good deal of research has gone into investigating the influence of the weather on violence. Apart from natural academic curiosity, the motivation seems to be to establish another parameter for judging crime figures. Is one year's improvement in murder figures, for example, a sign that policing policies are working better, or was it just because the sun was shining?
Here is a summary of some relevant academic papers:
"Weather and Crime" (Ellen G Cohn, British Journal of Criminology, Vol 30, 1990) examined 40 years of research linking climate and crime. "Assaults, burglary, collective violence, domestic violence, and rape tend to increase with ambient temperature, at least up to about 85F. The relationship between heat and homicide is uncertain." Her conclusion was that most violent crimes against persons increase linearly with heat, while property crimes are not strongly related to temperature change.
"The effect of temperature on crime" (Simon Field, British Journal of Criminology, Vol 32, 1992) reported that "temperature had a positive effect on most types of property and violent crimes". But no relationship was found between crime and either rainfall or hours of sunshine.
"The prediction of police calls for service: The influence of weather and temporal variables on rape and domestic violence" (Ellen G Cohn, Journal of Environmental Psychology, Vol 13, 1993) found that domestic violence seemed to be affected by the temperature and time of day, but that rape wasn't. Both domestic violence and rape happen more often after dark.
"The effects of weather on homicide" (Derral Cheatwood, Journal of Quantitative Criminology, Vol 11, 1995) found that "the consistent meteorological measure of importance is the number of previous days in a row in which the discomfort point (a specific point on the temperature and humidity index) has been over the physiologically relevant level of 79".
The overall picture of all this is rather confused. There seems to be more crime in summer, but none of the researchers consider the possibility that this could be because burglars don't like going out in the rain - or even because more people leave their houses unoccupied and prone to burglary in hot weather.
If a change in the weather does bring out the worst in our potential criminals, it could set in at a very early age, as the following result suggests:
"The relationship between weather and preschoolers' behaviour" (EL Essa, JM Hilton and CI Murray, Children's Environments Quarterly, Vol 7, 1990) found that two- to six-year-olds "interacted significantly more with peers and adults and less with materials during unstable weather or weather moving from unstable to stable" but pleasant, stable weather saw greater interaction with materials and less with humans. But we still need to know whether children beat each other up more when it's raining.