That was not I, I don't generally compare two cities. You can compare the same place over time(like the UK's murder rate climbing unaffected by stringent gun control laws, or similar occurances in US cities that have banned hand-guns). And also, a large number of relatively similar data points can help adjust for uncontrolled variables. That's the whole point behind many psychological and sociological surveys that use random sampling, a large enough number of data points shows a trend despite uncontrolled variables.