Nº 371 (April, 2026). Inés Berniell, Gabriel Facchini & Santiago M. Perez-Vincent

This study examines the challenges of analyzing domestic violence (DV) in Latin America and the Caribbean (LAC) and seeks to improve measurement through two main contributions. First, we collect and describe traditional DV data sources (house- hold surveys and administrative records) across 19 countries. The analysis reveals substantial gaps in data availability, with infrequent and outdated survey efforts in most countries. Nevertheless, surveys confirm high DV prevalence, with at least one in five women reporting victimization in all countries examined, and highlight perva- sive underreporting that limits the reliability of administrative crime data. Second, we examine the properties of a novel, high-frequency indicator based on online search behavior: the Google Domestic Violence Index. Using administrative data from eight LAC countries, we find that the index is strongly correlated with calls to DV helplines but shows weaker association with police reports or emergency calls. The evidence suggests that the index captures early-stage, information-seeking behavior and may provide a real-time signal of latent victimization not reflected in official statistics. Our findings underscore the potential of digital data to complement traditional sources and to support more timely, responsive approaches to tracking DV.

JEL codes: J12, J16, J18, I18

Suggested citation: Berniell, I., G. Facchini & S.M. Perez-Vincent (2026). Data Challenges and Innovations in Measuring Domestic Violence in Latin America and the Caribbean: Traditional Sources and Online Search Patterns. CEDLAS Working Papers Nº 371, April, 2026, CEDLAS-FCE-Universidad Nacional de La Plata.