How we measure neighborhood crime

We count crimes reported to police in small areas, adjust for how many people live there, and compare that to the rest of the country. The result is a rate you can compare between two places fairly, instead of a raw count that just tells you where the crowds are.

Why we use small areas

Most crime maps work at the city level. That is usually too coarse to be useful. A big city can hold both the safest and the least safe places in its whole state, and a single city-wide number hides all of it.

So we work at the level of a census block group. That is a small area, roughly neighborhood-sized, and it is the smallest geography the population data supports. You can see it for any address on the crime map.

Why we adjust for population

A downtown with 40,000 people will always report more incidents than a quiet block with 400. That does not make it more dangerous per person.

Dividing by the number of people who live there is what makes two places comparable. It is also why a busy commercial district can look worse than it feels: lots of people pass through, but few of them live there to be counted. We flag that where it matters.

Where the numbers come from

Two places. The FBI publishes national crime data that police departments submit to it. For some cities we also read the department’s own public records directly, which are usually more current and more detailed.

We prefer the more complete of the two for any given place, and we say which one we used.

The part most maps do not tell you

Police departments are not required to send their crime numbers to the FBI. It is voluntary. Most departments do it. Some do not. Some send only part of what they have, or stop for a year and start again.

This matters more than it sounds. A city can look safe simply because nobody turned in the data. If we drew that city anyway, we would be showing you a clean map of a place we know nothing about, which is worse than showing you nothing.

So we check each department’s reported volume against the FBI’s own published figures for the same area. Where the gap is too wide for the numbers to mean anything, we leave the area blank instead of filling it in. Blank means “we don’t know,” not “nothing happens here.”

What this cannot tell you

Worth being straight about the limits.

• It only counts crimes that were reported. Reporting rates differ between places and between types of crime, and nobody has good data on what goes unreported.
• It is historical. It describes what happened, not what will happen next.
• It says nothing about the people who live somewhere. It is a count of events in an area, and that is all it is.
• Small areas with very few residents produce noisy rates. One incident can swing the number a long way, so we treat thinly-populated areas with more caution than dense ones.

How often it updates

The FBI data lands on an annual cycle. The departments we read directly update far more often, some of them daily. Any page showing crime numbers carries the date we last checked it, at the top.

If you want to see this for your own area, the crime layer is on the map. It is free to use.