I would expect the crime levels to sort of coincide over the nation. In fact, they don't: http://www.wolframalpha.com/input/?i=violent+crime+in+seattle+vs+unit

I would expect the crime levels to sort of coincide over the nation. In fact, they don't: http://www.wolframalpha.com/input/?i=violent+crime+in+seattle+vs+united+states http://www.wolframalpha.com/input/?i=violent+crime+in+seattl... but it does seem that they do between these two cities. It is weird. Here's something else: http://www.wolframalpha.com/input/?i=violent+crime+in+washington+state+vs+seattle http://www.wolframalpha.com/input/?i=violent+crime+in+washin... http://www.wolframalpha.com/input/?i=violent+crime+in+washington+state+vs+united+states http://www.wolframalpha.com/input/?i=violent+crime+in+washin... Hypothesis: the country at large has a crime rate. In addition there is a magic variable which only affects cities, but affects all cities: http://www.wolframalpha.com/input/?i=violent+crime+in+san+francisco+los+angeles+chicago+vs+seattle http://www.wolframalpha.com/input/?i=violent+crime+in+san+fr...