Contact

GlossaryCrime reporting systems

National Crime Victimization Survey

The National Crime Victimization Survey measures specified nonfatal victimization, including events not reported to police. Learn its scope and limits.

The National Crime Victimization Survey, or NCVS, is a Bureau of Justice Statistics survey that measures specified criminal victimization from respondents' experiences. It includes events reported to police and events not reported to police. It complements law-enforcement records by providing a different source of information about crime.

What the survey covers

The NCVS interviews a nationally representative household sample, including eligible people age 12 and older. It collects information on specified nonfatal personal crimes and household property crimes. It is not a survey count of every crime category or every person in the United States. BJS survey description

For example, homicide cannot be measured by interviewing the deceased victim. The survey's household design also has coverage limits. These boundaries are reasons to identify the measured population and offense categories when describing a result, rather than call it a complete national crime count.

Survey weights allow responses from the sample to represent a broader population. Published estimates therefore differ from the raw number of interviews reporting victimization. Sampling uncertainty and methodological changes matter when comparing estimates; a small numerical difference does not automatically establish a meaningful change.

Example

Hypothetical example: A sampled respondent describes a qualifying theft that was never reported to police. The event can contribute to the survey's weighted estimate of victimization even though no corresponding police incident is available.

The survey has not created a missing police case file. It has measured an experience through a different collection method. Its national estimate also cannot be assigned to a specific small neighborhood merely because that neighborhood is part of the United States.

Common confusion

NCVS estimates and UCR police-recorded statistics are not competing versions of exactly the same measure. Their offense definitions, populations, information sources, and counting units differ. A disagreement between totals can reflect those design differences rather than an error in either source.

A victimization count is also different from the number of distinct victims. One person can experience more than one victimization. For household property crime, the household is the relevant unit, while personal victimization uses eligible persons. Rate denominators must follow those distinctions.

Reading a release

Use the methodology associated with the data year, including any redesign or comparability notes. Identify whether a result concerns all qualifying victimization or only the subset reported to police. Present confidence intervals or the publication's significance findings when assessing change. The survey's strength is evidence about covered experiences, including nonreporting; it should not be stretched into a precise local tally or a measure of offenses outside its scope.

Sources