Arrests of adults for domestic violence amount to about 60% of all adult arrest for interpersonal violence in the U.S. That estimate can reasonably be made from police data filed through the FBI’s National Incident-Based Reporting System (NIBRS). Special, state-defined domestic-violence reporting instruments are consistent with domestic violence arrest accounting for more than half of all arrests for interpersonal violence.
The Uniform Crime Reporting (UCR) system and the more extensive, more recently formulated National Incident-Based Reporting System (NIBRS) provide U.S. arrest statistics based on reports from local police jurisdictions. The Federal Bureau of Investigations (FBI) compiles statistics from local police reported under these systems in its annual Crime in the U.S. statistical collection. For crime incidents that police report, the FBI receives UCR reports from police jurisdictions covering about 93% of the U.S. population. The corresponding figure for NIBRS reports is roughly 27%. UCR data don’t provide incident detail to identify domestic violence arrests. NIBRS, in contrast, includes sufficient detail to identify reported incidents, victims, offenders, and arrests for domestic violence.
NIBRS statistics describe domestic violence from the perspective of the police. They do not cover domestic violence that does not come to the attention of the police. Moreover, the terms “victim” and “offender” reflect police judgment of the incident. These terms do not indicate a criminal justice system judgment under law that a domestic violence crime has occurred and that a particular person is guilty of committing domestic violence against another person. From the perspective of justice under law, implicit in NIBRS (and UCR) categories of incidents and persons is the qualifier “alleged.”
Extrapolating NIBRS statistics to a nation-wide estimate requires considering jurisdiction population size non-reporting bias. NIBRS coverage of persons in cities and places with population equal or greater than 250,000 persons (reporting agencies cover 15% of the total U.S. population) is about half that of coverage cities and places with population less than 250,000 (29% reporting agency coverage). The three largest cities in the U.S. — New York, Los Angeles, and Chicago — do not file NIBRS reports. The largest jurisdiction reported under NIBRS is Fairfax County, Virginia. Fairfax has a covered population of about a million. A table of NIBRS-reporting jurisdictions with population covered greater than 100,000 is available in the punishment-us-dv-arrests workbook. If a NIBRS statistic varies significantly across population jurisdiction size, then population size bias in non-reporting is significant for forming a national estimate from NIBRS.
The relational scope of domestic violence is defined here using the definition from a relatively high-quality Virginia report about domestic violence measured by NIBRS statistics. The offenses used to define violence are those recorded for up to three victims in the associated incident, not the single arrest offense code. If an arrest is associated with any domestic violence victimization, it is counted as a domestic violence offense. Since most domestic violence arrests are coded as having one offender and one victim, the counting approach to multiple offender / multiple offense arrests incidents isn’t likely to have a large effect on the domestic violence arrest share.
NIBRS statistics indicate that domestic violence arrests account for 54% of all arrests for interpersonal violence. The corresponding figure among adults arrested is 58%. Interpersonal violence comprises the NIBR arrest offense categories homicide offenses, forcible sex offenses, robbery, assault offenses (including simple assault and intimidation), and kidnapping. The domestic violence arrest shares differ little between large cities and all other cities providing NIBRS data. They are also consistent with state-specific figures for states having nearly 100% NIBRS reporting and with independently compiled state figures. The domestic violence arrest share excludes in both its numerator and denominator interpersonal violence reported as “disorderly conduct” in a Part B NIBRS arrest report. Applying the domestic violence arrest share to a UCR-based estimate for all arrests for interpersonal violence (including simple assault) implies about a million arrests for domestic violence across the U.S. in 2010.
Arrest reporting appears to be less complete for arrests for less serious offenses. Arrests for interpersonal violence account for 34% of all arrests in NIBRS statistics. The corresponding figure for UCR arrests statistics is 14%. For Iowa, state-published data suggests that that Iowa District Court criminal cases filed for “domestic abuse assault simple misdemeanors” are not reported under incident-based reporting as domestic violence arrests. More generally, national statistics suggests that commitments to lockups and jails are significantly higher than UCR reported arrests. A plausible explanation for these statistical patterns is that police officers allocate less reporting time for less serious offenses. Hence NIBRS provides a less comprehensive view of all arrests than does UCR, and UCR a less comprehensive view of all arrests than lockup and jail commitment statistics.
Persons in jail for a family violence conviction account for a much smaller share of persons in jail serving any type of conviction for violence. Statistics for 2003 indicate that 22% of convicted violent offenders in jail have been convicted for family violence. The definition of family violence used in calculating that statistic is “family violence includes all types of violent crime committed by an offender who is related to the victim either biologically or legally through marriage or adoption.”^ Domestic violence is defined broadly under law. Compared to common legal definitions of domestic violence, this statistical definition of family violence thus excludes violence between ex-spouses, violence between cohabiting intimate partners, and violence between other current and former boyfriends and girlfriends. Including jail inmates convicted for violence against (current) boyfriend or girlfriend with jail inmates convicted for family violence gives a 42% share of convicted violent offenders. As a result of domestic-violence mandatory arrest laws, persons arrested for domestic violence are likely arrested for much less serious offenses than persons arrested for non-domestic violence offenses. That selection effect makes the share of domestic-violence arrests among all arrests for violence greater than the share of jail inmates convicted for domestic violence among all jail inmates convicted of violence.
Arrest is a common experience across young men. Based on data from the National Longitudinal Survey of Youth 1997 (covering the years 1997 to 2008), by age 23, 40% of men have been arrested at least once. The corresponding figure for women is 19%.^ The above arrest shares for {domestic violence / all interpersonal violence} and {all interpersonal violence / all arrests (excluding traffic violations)} suggest that roughly 3% of men have been arrested for domestic violence by age 23.
Related datasets: