I often think per-capita statistics are misleading …

1. The bias of averaging. Relatively few people in a community is likely to have an “average” experience to a problem. Often it’s handful of people contributing to the extremes, and even those closer to the middle are likely to have smaller or larger impact the statistic might suggest. Crime might be really bad in neighborhood but other neighborhoods are quite safe.
2. The bias of noise in rural communities. Any community with a relative small population is likely to be biased by just random chance, and is not reflective of typical events happening on ground. One  random event in a community of 1,000 people is going to look a lot worse on paper then ten random  events in a community of 100,000.
3. The bias of smoothing out the mean in urban communities. Large urban communities on paper often appear to have lower per-capita emissions or acts of violence. That’s because while there may be some randomness in data, there are just so many more people to smooth out biases in the data causing randomness.
4. Not recognizing that one community is connected to another. Often a factory, a farm or other rural business can be producing products for an urban community, so emissions that appear on a chart for one community, might actually be attributable to another community through the consumption of their product. For example, you see this with China being the industrial exporter of the world. Or a per-capita crash statistic recorded in a rural area, even though the motorist was passing through between cities and didn’t really impact the local area.
5. Total emissions or crimes might be within the local community’s ability to manage them. Pollution is after all emissions at a level that cause negative impacts to an environment, at levels that existing ecological services can’t absorb. Likewise, often big cities are better equipped to deal with crime, and while overall crime rates might be higher, the impact is lower to most individuals then we might think due to effective policing.

I think it’s a mistake to use per capita statistics to stigmatize or lay blame on a community. Often problems that raise to the level of community concern involve the entire community, not a single area that is easy to point fingers too, especially if it’s not our own. Dividing a total population into individual heads by math can be misleading and must be used with care. While we all like a colorful map, think about what you seeing in the map, and people it represents before coming to conclusions based on a mathematical model.