FBI Crime Data Is Public — Here's How to Actually Use It

Every time someone tells me a neighborhood is "dangerous," I ask the same question: compared to what?

The FBI has that answer. It's free, it's public, and it's sitting online right now at a URL most people have never heard of. What it doesn't come with is a plain-English guide to actually using it — which is what this piece is.

UCR vs. NIBRS — What's the Difference and Why It Matters

The FBI collects crime data through two parallel systems, and understanding the difference explains why two websites can show different numbers for the same city.

The Uniform Crime Reporting program (UCR) is the older system, dating to 1929. It works by collecting summary counts from participating agencies: how many robberies, how many thefts, how many arsons. It tells you totals. It doesn't tell you much about the individual incidents.

The National Incident-Based Reporting System (NIBRS) is the newer, more detailed replacement. Instead of summary counts, NIBRS captures structured data on each individual incident: what type of offense, what time, what location type (residence, parking lot, street), the relationship between victim and offender, and more. More agencies are transitioning to NIBRS every year, but coverage is still incomplete — and a city on UCR gives you less to work with than a city on NIBRS.

This is why the number on one website might not match another: they might be using different reporting systems, or one site might be filling in missing agency data with estimates while another leaves gaps.

How to Find Your City's Data

The FBI's Crime Data Explorer is the authoritative public source: crime-data-explorer.fr.cloud.gov. It's not the prettiest interface ever built, but it works.

The basic flow:

1. Go to the Explorer and click "Agency" to look up a specific police department (most useful for city-level data).
2. Select the offense category you care about. "Property Crime" and "Violent Crime" are the main top-level buckets. You can go more specific — burglary, theft, motor vehicle theft, assault, robbery.
3. Select a year range. I recommend at minimum three years to identify a trend.
4. Note the population figure the FBI is using. This matters for your rate calculation.

What you'll get is raw counts. Do not stop there.

The Math Nobody Does (But Should)

Raw count ÷ population × 1,000 = rate per 1,000 residents. That's it. That's the formula.

Without this step, you are comparing numbers that cannot be meaningfully compared. A city of 2 million with 2,000 burglaries has a rate of 1.0 per 1,000. A small city of 50,000 with 500 burglaries has a rate of 10.0 per 1,000 — ten times worse, even though the raw number is smaller.

This mistake is so common that it's probably responsible for a majority of neighborhood misconceptions. Big-city crime numbers look huge in absolute terms while being perfectly manageable at the rate level. Small-town crime numbers look small while sometimes representing serious density when normalized.

The FBI data is the floor of what's happening — not the ceiling. The real number is always higher. That's not a reason to panic. It's a reason to contextualize.

— Maya Reyes

What the FBI Data Doesn't Tell You

Know the limits before you trust the numbers:

• Underreporting is real and significant. The Bureau of Justice Statistics runs a separate survey (the National Crime Victimization Survey) that captures crimes regardless of whether they were reported to police. The gap between the two datasets suggests 40–60% of property crimes and a significant share of violent crimes are never reported. FBI data reflects crimes known to police — not all crimes.
• Some agencies don't report or report inconsistently. A city with missing years or partial coverage will show anomalies. Check whether coverage is complete for the agencies in your area.
• The data is always at least 12–18 months behind. The FBI publishes annually, and there's always a lag. You're working with historical data, not real-time intelligence.
• Geographic unit matters. City data covers city limits. County data covers a wider area that may include much safer rural zones, which pulls the average down. Metro area data is different still. Make sure you're looking at the right geographic unit for what you're evaluating.

We've built a translation layer on top of this public data — pulling from FBI, NIBRS, local police feeds, and state DOT records — so you don't have to do this manually for every address you're considering. But now you know where it all comes from. That's not nothing.