The Armed Conflict Location & Event Data Project (ACLED)¹ is one of a number of recently developed high disaggregated armed conflict datasets. Similar to The Uppsala Conflict Data Program Georeferenced Event Dataset (UCDP-GED v18.1),² The Social Conflict Analysis in Africa Database (SCAD),³ and the Militarized Interstate Dispute Locations (MIDL v4.3)⁴ datasets, ACLED presents dyadic conflict data with a high level of spatial accuracy. Although these datasets were initially released with restricted geographic scope (with the exception of MIDL), they all expanded their spatial reach significantly over the past 5-10 years. ACLED now encompasses Africa, South Asia, South East Asia, the Middle East, Europe, and Latin America, UCDP-GED is currently advertised as a global dataset of armed conflict, and SCAD has expanded to include Africa, Mexico, Central America, and the Caribbean.

These datasets are similar in nature, but can vary greatly in their reporting criteria and supplementary information. Eck (2012)⁵ recently reviewed the fundamental differences between UCDP-GED and ACLED. UCDP-GED provides superior data for violence and fatalities, but ACLED distinguishes itself by being the only source for non-violent conflict related events like troop movements and governmental aid. Furthermore, in contrast to UCDP-GED, ACLED uses less restrictive definitions of “events” and “actors” involved in an event. The effect of non-violent event coding and looser actor definitions is a larger database, however, it may be less suitable for statistical inference and analysis without substantial pre-processing. Spatially disaggregated conflict data presents additional challenges. Larger conflicts between the same actors often result in several battle events that are recorded as individual observations. Although UCDP-GED provides several metadata fields and additional datasets to better characterize the complexities of inter and international conflict, ACLED provides limited metadata and no additional datasets.