The Visibility Deficit Index: Triangulating Nighttime Lights and Event Data to Overcome Conflict Illegibility

With Jorge Ruiz Reyes.

Can we systematically identify places where violence is likely to go underreported? Conflict data are shaped by the same conditions that render violence difficult to observe, producing non-random patterns of incomplete documentation. We propose a triangulation strategy that combines ACLED with satellite-derived nighttime lights (NTL) to identify areas of conflict invisibility. Our core contribution is the Visibility Deficit Index (VDI), which measures the gap between NTL-detected disruption and reported conflict events. We apply this framework to cartel violence in Mexico, focusing on Michoacán since 2021. Results reveal a stark territorial gradient: municipalities under cartel control exhibit high visibility deficits, indicating severe underreporting. By leveraging divergence between data sources, the VDI provides a scalable tool to detect hidden violence and improve the legibility of conflict in otherwise opaque settings.