Estimating Domestic Violence Trends: Addressing Underreporting for More Accurate Crime Data

• In Event: Data, Policy, and Experience: Justice-Involved Women and Domestic Violence

Abstract

Domestic violence remains one of the most difficult crimes to measure accurately due to inconsistent definitions, underreporting, and fragmented reporting mechanisms across agencies. While law enforcement data from the National Incident-Based Reporting System (NIBRS) provide official counts of domestic violence incidents, they fail to capture the true prevalence of victimization. Similarly, while the National Crime Victimization Survey (NCVS) offers self-reported victimization rates, it does not allow for jurisdiction-level estimates. This study presents a novel methodology that combines NIBRS and NCVS data to generate more accurate, sub-national estimates of aggravated domestic violence (link to code provided).

Findings indicate that official reports undercount aggravated domestic violence incidents by approximately 29% to 53% in major U.S. cities. By adjusting reported incidents using self-reported victimization data and demographic factors, this study provides a more accurate depiction of domestic violence prevalence. The methodology is scalable, allowing local agencies to estimate underreported crime in near real-time. Additionally, this approach is applicable to other crimes prone to underreporting, such as hate crimes.

Accurate domestic violence data are crucial for resource allocation, policymaking, and intervention strategies. This study underscores the importance of accounting for underreporting to enhance public safety responses and improve victim support services.

Authors

• Alex R. Piquero, University of Miami

• Andrew Wheeler, CRIME De-Coder

• Stephanie C. Kennedy, Council on Criminal Justice