In The Unheavenly City Revisited (1974), Edward Banfield notes how many urban problems only affect parts of cities:
Crime is a partial exception, but in Chicago (so the Violence Commission was told) a person who lives in the inner city faces a yearly risk of 1 in 77 of being assaulted whereas for those who live in the better areas of the city the risk is only 1 in 2,000 and for those who live in the rich suburbs only 1 in 10,000.
So, things are one order of magnitude safer in the suburbs than the city, and one order of magnitude worse in the inner city than in the good parts. I suppose that sounds about right.