The Crime Survey That Counts Unreported Crime (NCVS)

Last updated October 29, 2026

Arguments about whether crime is going up or down often rely on one of two kinds of data: crimes reported to police, or crimes people say happened to them in a survey. The survey is the National Crime Victimization Survey (NCVS), run by the Bureau of Justice Statistics (BJS). It matters because, by design, it includes crimes that were never reported to police.

What the NCVS is

BJS calls the NCVS the nation’s primary source of information on criminal victimization. Each year, it gets data from a nationally representative sample of about 240,000 persons in about 150,000 households. People are interviewed about the frequency, characteristics and consequences of criminal victimization in the United States.

Which crimes it covers

According to BJS, the NCVS collects information on nonfatal personal crimes (rape or sexual assault, robbery, aggravated and simple assault, and personal larceny) and household property crimes (burglary/trespassing, motor vehicle theft and other theft), both reported and not reported to police. BJS lists only nonfatal personal crimes, so homicide isn’t part of it.

What else it records

BJS says respondents provide information about themselves, such as age, sex, race and Hispanic origin, marital status, education and income, and whether they experienced a victimization. For each incident, the survey collects details about the offender (including age, sex and victim-offender relationship) and the crime (time and place, whether weapons were used, the nature of any injury, and more).

Why survey and police numbers can disagree

Unreported crime. The NCVS includes incidents victims never reported, which police-reported data can’t capture.
Different crimes. The NCVS covers only the nonfatal personal and household property crimes listed above; other datasets may cover a different set.
Different methods. One is a sample survey with estimates and margins of error; police-reported data can only include incidents police know about.

Two statements, “violent crime fell” and “victimization didn’t fall,” can both be accurate if they cite different measures or years. Ask which dataset and which crimes.

Checklist for crime claims

Is the source a victimization survey or police-reported data?
Which crimes are included?
Which year, and compared with what?
A rate (per population) or a raw count?

See our worked example: “Violent crime went down in 2024.”

What the survey asks about each incident

Beyond who was involved, BJS says the NCVS records economic consequences to the victim, whether the crime was reported to police, the reasons it was or wasn’t reported, and the victim’s experiences with the criminal justice system. That makes it one of the few sources that can say anything about why crimes go unreported.

A long-running series, with changes to note

BJS says the survey has been ongoing since 1973. BJS also lists an NCVS instrument redesign, including full implementation of a new instrument, and a sample redesign. When comparing years, check BJS notes for any changes in methods. BJS publishes supplements such as the Identity Theft Supplement and the Police-Public Contact Survey, and a dashboard tool (N-DASH) for exploring the data. Recent publications listed on the program page include preliminary victimization estimates for 2025.

Sources

Bureau of Justice Statistics: National Crime Victimization Survey. Checked October 29, 2026.

More explainers

Last updated October 29, 2026. Facts checked October 29, 2026 against the official sources linked above. EvenFacts is independent and isn’t affiliated with any government agency, party or candidate. This explainer is general information, not legal or financial advice, and it doesn’t take a side.