About CrimeContext

Why FBI UCR data deserves an honest public-facing home

The FBI maintains the Uniform Crime Reporting (UCR) program and its successor, the National Incident-Based Reporting System (NIBRS). Together they constitute the federal record of every crime reported to police by every participating law-enforcement agency in the United States. The data is the only nationally consistent source for comparing crime rates across cities and counties — and it is also one of the most frequently miscited datasets in journalism, because the reporting rules, the agencies, and the data quality vary substantially across jurisdictions.

CrimeContext is built to present UCR/NIBRS data without the most common analytical mistakes. Every city and county page shows the per-100k rates for violent crime and property crime alongside the multi-year baseline, with explicit notes when the reporting agency missed years or transitioned from UCR to NIBRS in a way that breaks the historical comparison. The data is the FBI’s; the value the site adds is honest framing.

How the pipeline pulls FBI data

The pipeline pulls from the FBI Crime Data Explorer agency-level NIBRS and Summary Reporting System files on the FBI’s semi-annual release cadence. Full-year data typically arrives in the September release of the following year. The pages stamp the as-of date and the reporting-agency name on every figure, so a reader can verify any value against the FBI source.

A practical detail: the FBI’s transition from UCR to NIBRS, which most large agencies completed by 2021, produced a real break in the historical series. The pages flag affected years and avoid comparing pre- and post-transition years for the same agency without an explicit note.

Where FBI crime data has caveats

Three things to know. First, not every agency reports every year. Voluntary participation in UCR and NIBRS means that gaps in the data look like zeros if read naively, but they are often missing reports rather than missing crime. The site flags agency-level reporting gaps explicitly.

Second, FBI UCR captures reported crime, not actual crime. The reporting rate varies by offense (property crime is underreported, violent crime less so) and by jurisdiction (depending on community-police trust). Comparing two cities’ reported rates without that context can produce misleading conclusions.

Third, UCR/NIBRS excludes federal, tribal, and university agencies unless they voluntarily submit. Counties with significant federal-jurisdiction land or tribal-land coverage will show artificially low rates The methodology page on this site documents every dataset, every refresh cadence, and every limitation in detail so readers can trace any numeric value on the site back to the underlying federal source. We treat that traceability as a hard requirement for any data product that asks readers to make real-world decisions on its output. because the federal and tribal incidents are not in the state’s UCR/NIBRS submission.