Periodic Reporting for period 1 - ANTICIPATE (Anticipating the impact of armed conflict on human development)
Reporting period: 2022-12-01 to 2025-05-31
Armed conflict is human development in reverse. The full scale of conflicts’ impacts remains unknown, however, and fragmentation of research into multiple academic fields limits our understanding. This multi-disciplinary project brings together scholars from economics, political science, and conflict research to study the impacts in much more detail and comprehensiveness than earlier studies. It takes a risk-analysis perspective, assessing the expected impact as a function of hazard, exposure, and vulnerability, and consider effects at both the macro and micro level, on economies, human development, and political institutions. It will model exposure to conflict events by accounting for how effects of observed, overt violence are transmitted to locations far from the violence itself and over time, identify conditions that make local communities, marginalized groups, and women particularly vulnerable to the effects, and study how conflict increases their vulnerability to other shocks such as natural disasters. The objective of the project is to model hazard as a probability distribution over the predicted number of direct deaths from violence in locations across the world, exposure as a model for the extent to which local populations are affected by this likely violence, and vulnerability as how exposure is translated into adverse human development impact for these populations. The results will be coordinated in the form of a monthly updated early warning system, expanding the well-established ViEWS model, to also alert observers to particularly detrimental occurrences of violence. Throughout, the project will study how the various impacts and vulnerabilities identified work to reinforce each other, and formulate policy recommendations for parties seeking to reduce the impact on human development.
As stressed in the influential report Pathways to Peace developed jointly by the UN and the World Bank forceful action in the face of conflict-induced disasters requires that we fully appreciate and anticipate the extent of the havoc. Early warning allows early action. The core aim of ANTICIPATE is to contribute significantly to assisting stakeholders in anticipating the negative humanitarian impact of armed conflict, in the form of a sound theoretical and methodological basis for a systematic, quantitative impact forecasting system. To be useful, such a system must be transparent, systematic, have uniform coverage across the geographical area it covers, and be updated frequently. It should look a few years into the future, a time range for which action could plausibly be taken and have effect. It must be possible to link the anticipated impact to the political violence that triggers it. Then, even when politics – violent and non-violent – render humanitarian assistance infeasible, the warnings can be used to demonstrate the adverse consequences of the choices made by cynical actors.
As stressed in the influential report Pathways to Peace developed jointly by the UN and the World Bank forceful action in the face of conflict-induced disasters requires that we fully appreciate and anticipate the extent of the havoc. Early warning allows early action. The core aim of ANTICIPATE is to contribute significantly to assisting stakeholders in anticipating the negative humanitarian impact of armed conflict, in the form of a sound theoretical and methodological basis for a systematic, quantitative impact forecasting system. To be useful, such a system must be transparent, systematic, have uniform coverage across the geographical area it covers, and be updated frequently. It should look a few years into the future, a time range for which action could plausibly be taken and have effect. It must be possible to link the anticipated impact to the political violence that triggers it. Then, even when politics – violent and non-violent – render humanitarian assistance infeasible, the warnings can be used to demonstrate the adverse consequences of the choices made by cynical actors.
The research has focused on the methodological foundation for the research, a core task in the proposed project, with several studies exploring how to best include uncertainty in armed conflict and impacts forecasting models, how to evaluate such models, how to deal with the uncertainties of the input data, and how to handle the challenging distribution of the outcomes under study. The project has also focused on the study of the specific impacts of armed conflict through single-outcome studies and literature reviews. The project has laid the ground for the joint modeling and the early-warning system by developing an advanced infrastructure to enter data into the project database with a battery of validation tests to ensure data integriry, to extract it in useful fashion, and on developing a ground-breaking, advanced machine-learning operations pipeline.
The project is well under way to construct a full representation of uncertainty, all the way from incompleteness and missingness in input data, through statistical uncertainty, and specification uncertainty, to inform the uncertainty of the forecasts of armed conflict as well as its humanitarian impacts. This has not been done in this field before.
Several of the models derived in the project also go beyond the state of the art, e.g. the novel Artificial Neural Network model, models adapted to the zero-inflated extreme-value distribution of the outcome data.
The project has involved a set of leading conflict forecasting environments globally in an ongoing, friendly prediction challenge, the results of which will be highly informative when it is finalised.
Several of the models derived in the project also go beyond the state of the art, e.g. the novel Artificial Neural Network model, models adapted to the zero-inflated extreme-value distribution of the outcome data.
The project has involved a set of leading conflict forecasting environments globally in an ongoing, friendly prediction challenge, the results of which will be highly informative when it is finalised.