Is there an impact of weather/season and the amount of crime happens – The Jury is out on that (see here,here and here). Clearly the issue has been studied quite a bit (this from a paper from 1952) and this one more recently with no conclusive result just yet.
The most commonly seen result (so far) is that precipitation does not have a significant correlation with crime but temperature apparently does. Here is a nice summary of the research about this topic. I doubt if such a correlation would hold in an Indian Context where the amount of economic activity is significantly dependent on how the amount of rainfall a particular region gets in the season.
Anyway, while researching for this earlier post , I landed on a crime stats dataset for Bihar that had data split by different types of crime per month for the last years.Once I started I digging into the data initially for murders (since that was the topic of the previous post)a curious observation jumped out from the graphs. For some reason the number of murders seemed to be always higher in months between May and July for every year.
Intrigued as I was, I checked this trend for other violent crime categories listed by the Bihar Police. Here is what I find.
Notice those huge bumps in the middle of the year.Only one category of crime (as listed in the data set) did not show this behavior.
I am not sure if these trends mean anything in a behavioral sense i.e. people don’t get extra violent in summer or grow more docile in the winter. If any thing, these trends might just reflect some underlying economic patterns/hardships which perhaps manifested in increased crime.