If the surveying is accurate and representative, it's useful because it means that crime is unusually indiscriminate (high rates of anonymous crime.) Crime mor

If the surveying is accurate and representative, it's useful because it means that crime is unusually indiscriminate (high rates of anonymous crime.) Crime more typically happens between people with connections to each other. For example, my city has a per capita crime rate similar to San Fransisco's, but most don't experience it unless they live or work in a few neighborhoods (or single intersections in some cases.) In my city, something like 5-10% of people would report having recently been a victim of a crime, despite the number of crimes per capita being similar.