The overall ACE project is motivated by the need to better understand the behaviour of non-state armed groups in perpetrating atrocity crimes such as crimes against humanity, ethnic cleansing and war crimes. The data collection is based on six countries Central African Republic, the Democratic Republic of Congo, Iraq, Nigeria, Syria, and Somalia with a focus on non-state actor perpetrated atrocity events. The aim is to have a fine-grained event data collection of different types of atrocity crimes in these countries. These event types are derived from the Rome Statute. More specifically, the unit of observation in ACE is the event where a named state or non-state actor is involved on a given day in a specific location. Each individual event is covered with the best precision regarding the type of event, location, perpetrator and victims.Since 2010, there has been a 'dramatic resurgence' of violent conflict around the world (United Nations, 2018, p. v). As part of this trend, mass atrocity crimes, defined as genocide, war crimes, crimes against humanity, and ethnic cleansing (GWCE), have become 'the new normal' (Human Rights Watch 2018). At this time of writing, the Global Centre for the Responsibility to Protect (GCR2P) identifies seven countries that are 'currently' experiencing GWCE, three at 'imminent risk', seven of 'serious concern', and thirteen being 'monitored' because they have experienced GWCE in the recent past (Global Centre for the Responsibility to Protect 2019). These crises have seen millions of people killed, tens of thousands raped, and underpin an unprecedented refugee crisis. Although mass violence is not a new phenomenon, non-state armed groups such as Al Qaeda, Islamic State, Boko Haram, Lord's Resistance Army, and Al-Shabaab are increasingly playing a critical role in the perpetration of atrocity crimes leading to key policymakers calling for urgent research on this specific threat (see case for support). Responding to this new reality, the project answers the following primary research question: under what conditions do non-state armed groups perpetrate GWCE? The funding will enable us to develop the first dataset in the world that collects systematic evidence on non-state armed groups perpetrating GWCE, which we call 'Atrocity Crime Events' (ACE) dataset. To do this, we will analyse six countries and three themes. The former refers to the Central African Republic, the Democratic Republic of Congo, Iraq, Nigeria, Syria and Somalia. The latter focuses on i) interactions, for example, between the non-state armed group[s] themselves, other actors such as the government, and external actors such as UN peacekeepers, ii) local factors, for instance, geography, economics, population density, as well as natural resources, and iii) group characteristics, such as age, ideology, and external support. The scientific impact develops in three stages. First, the unique dataset 'ACE' will provide the necessary information to run statistical analysis to explain why, when, and where mass atrocities happen in our six chosen countries. Second, we will develop hypothesis based on our three themes that can be tested through future academic inquiry. Third, the project seeks to drive forward quantitative research into the causes of non-state armed groups perpetrating mass violence. This advance in knowledge will allow us to provide policy recommendations in order to improve international, regional, and national strategies toward mass atrocity prevention with a specific focus on policymakers in the United Nations (UN), the European Union (EU), the United Kingdom (UK), and Africa (the four case study governments and organisations such as the African Union). We will work with three project partners, GCR2P (New York and Geneva), Aegis Trust (Kigali), and Protection Approaches (London), as well as an advisory board consisting of Alex Bellamy, Gyorgy Tatar, Ivan Simonovic, Karen E. Smith, and Kristian Skrede Gleditsch. As part of our impact strategy, we will hold end of project workshops in London, New York, and Kigali. Outputs will include i) publicly available dataset and codebook, ii) six articles in high ranking journals, iii) an Analysis Framework for the United Nations Office on Genocide Prevention and the RtoP, iv) co-created policy reports with each project partner focusing on the UN, the UK, the EU, and African mass atrocity prevention strategies, v) blogposts, vi) infographics, and vii) presentations at conferences and policy-orientated meetings.

The data collection methodology is based on coding news reports extracted from LexisNexis. The extraction of news reports from LexisNexis has been narrowed down by using specific search terms for each event type, including the countries in this project. The focus is primarily on English language sources and where necessary, the geography filter is used to narrow down results based on the location of the event. Once a set of news reports have been identified from Lexis Nexis, the coders skim through the reports based on headlines/short descriptions and select to read through the ones that seem to constitute an event (as opposed to, for example, reports about UN meetings to discuss atrocities etc.). The coders then write a short description of the event on the dataset and code the rest of the variables in the dataset with best precision possible. The coding of the events is based on ACE codebook and is conducted by human coders, each specialising in one of the countries in question.