Thats not how statistics work. The average number of crimes per person is going to be the same so the crime per capita is the same as the amount of crime an person should expect to experience on average.
You may be thinking of a median or mode, which can differ based on the shape of the distribution curve.