The Political Declaration makes clear that data collection is not just about recording the number of civilians killed and injured by explosive weapons, but also on the reverberating and longer-term impacts. Data collection methodologies, which are largely orientated towards monitoring the direct impacts of explosive weapon use, must be considered alongside this broadened understanding of harm to civilians. This article provides insight into current data collection efforts, reflects on questions raised by the Explosive Weapons Monitor as it works to strengthen its reporting of civilian harm, and identifies ways in which this reflection might contribute to the development of norms and good practices in tracking the full scope of harm to civilians from the use of explosive weapons.
Introduction
The Political Declaration on Strengthening the Protection of Civilians from the Humanitarian Consequences Arising from the Use of Explosive Weapons in Populated Areas represents a broadened understanding of the scope of harm to civilians from the use of explosive weapons and signifies that data should be collected on both the direct and indirect, or reverberating, effects of the use of explosive weapons.
In doing so, the Declaration makes clear that data collection is not just about recording the number of civilians killed and injured by explosive weapons, but on the reverberating and longer-term impacts as well, such as damage to schools and hospitals and other civilian infrastructure, the disruption of education and healthcare and other essential services, and psychological trauma.
This provides a more comprehensive understanding of the impact of explosive weapons use on civilians and provides an evidential basis for harm reduction that can inform operational changes and responses. It also provides important information about the progress made in fulfilling the humanitarian commitments of the Political Declaration and in avoiding the use of explosive weapons in populated areas.
Data collection methods and methodologies, which are largely orientated towards monitoring the direct impacts of explosive weapon use – impacts that result from an identified incident of explosive weapon use at a specific time and location – must be considered alongside this broadened understanding of harm to civilians. Indirect effects of explosive weapons use – impacts that occur as a result of incidents of explosive weapons use that cause civilian harm beyond the time and location of an attack – are generally more challenging to document as a result of this and other factors.
This article provides insight into the current landscape of data collection by research partners of the Explosive Weapons Monitor which provide a conceptual basis for understanding harm to civilians from the use of explosive weapons. It reflects on questions raised by the Explosive Weapons Monitor as it works to strengthen its reporting of civilian harm from explosive weapons use and addresses challenges in this documentation. Finally, it identifies ways in which this reflection might contribute to the development of norms and good practices in tracking the full scope of harm to civilians from the use of explosive weapons.
Decades of data collection on civilian harm from the use of explosive weapons
The Explosive weapons Monitor sources data from existing datasets managed by Action on Armed Violence (AOAV) and Insecurity Insight, both of which utilise methodologies derived from decades of efforts to monitor the impacts of weapons and contribute much-needed evidence to support advocacy on the need for action to address the issue of explosive weapons in populated areas.
Data that shows the need to prevent the harm caused by weapons was essential to the International Committee of the Red Cross (ICRC) Superfluous Injury or Unnecessary Suffering (SIrUS) project, initiated in 1996, which sought to determine how specific categories of weapons foreseeably cause military casualties.1 With this endeavour, the SIrUS team, including surgeon Robin Coupland, established a robust precedent for expressing evidence-based concerns by taking available data from recent conflicts into account.2
Applying a methodology developed by Coupland and Nathan Taback, an academic statistician, Landmine Action generated a dataset on casualties caused by explosive weapons use worldwide that included data recorded from media reports in English-language newswire sources. Landmine Action is now AOAV and since 2010 has provided continuous monitoring of explosive violence on that basis. This includes data disaggregated according to casualty type (deaths or injuries), civilian status (civilian or armed-actor), and location (populated or unpopulated), amongst other indicators.
In the early 2000s, Coupland, Taback, and Christina Wille, developed a methodology for ‘documenting, presenting and using data on violence’, which developed into Insecurity Insight. It supported healthcare in danger research before expanding thematically to other areas of concern over humanitarian access in conflict.3 This method, which identifies the potency of harm, as well as the vulnerability to such harm, underlines the data collection undertaken and shared with the Explosive Weapons Monitor on access to aid, education and health care.4
Challenges to obtaining consistent, representative data
Persistent efforts by civil society and international organizations continue to improve data collection and analysis on the humanitarian consequences of the use of explosive weapons in populated areas. Yet, challenges from earlier identified methodologies persist, along with new and added challenges for monitoring the indirect or reverberating effects of explosive weapon use.
Under-reporting
Landmine Action succinctly summarised key limitations of data in 2009, based on the experience of its initial monitoring of incidents of explosive weapons use:
“The data are not comprehensive and are subject to limitations and qualifications – these are primarily general under-reporting and geographical biases; the data are also likely to under-report combatant casualties and civilian casualties from major military engagements.”5
Examples of challenges then, and now, that contribute to under-reporting include:
- Lack of access to information during peak conflict periods
- Uneven and shifting media attention on specific conflicts
- Constraints in language capacities for interpretation of local media and social media
- Conflict reporting sources which have a limited focus on fatalities (for example, sources reporting within the frame of Sustainable Development Goal 16.1.2)
- Data sources that record information on survivors that are able to reach medical care (for example, data gathered from emergency surgery wards and hospitals)
- Inconsistencies in the ways in which data is reported (for example, some sources regularly gather and disseminate immediate incident or casualty reports and others collect and release aggregate casualty or incident numbers for a fixed period)
Documenting reverberating effects
Reverberating, or indirect, effects of explosive weapons use – impacts that occur as a result of incidents of explosive weapons use that cause civilian harm beyond the time and location of an attack – are generally more challenging to document.
The damage that explosive weapons cause to civilian infrastructure are known to have long-lasting consequences, displacement, disruption or disintegration of essential services, and lack of access to medical care. These impacts, combined in any which way, may result in further disease outbreaks and other public health emergencies, food insecurity and economic crises.
The harm and deprivation of reverberating effects on civilians are broadly apparent in retrospect. Yet the direct evidence of these effects stemming from an incident of explosive weapon use can be difficult to glean while they are occurring. This is due in part to a lack of baseline information, as well as the very interconnected nature of the process and systems.
Overcoming challenges and identifying good practice
Much has been done to overcome these challenges and to work towards strengthened research and analysis on the full scope of harm to civilians from the use of explosive weapons. In its own efforts to strengthen its reporting of civilian harm from explosive weapons use and address challenges in this documentation, the Explosive Weapons Monitor has considered various means and methods for increasing the scope of its reporting.
Identifying new sources of data
New and/or under-explored sources of relevant data can help increase our understanding of the impacts and humanitarian consequences on civilians from the use of explosive weapons. A number of emerging, diverse sources are now increasingly available. These include:
- Geospatial digital data
- Open-source intelligence (OSINT) investigative groups
- ‘Before and after’ satellite imagery and satellite measurements of airborne substances and pollutants
- Social media content analysis
- Qualitative reporting, such as field investigations, first-person interviews and surveys
Each source offers unique opportunities for measuring harm from explosive weapons, though may employ differing methodologies in order to do so.
Incorporating key concepts into data collection systems
A small number of sources on civilian harm have been specifically designed to record the impact of explosive weapons in populated areas. Several other sources also record data disaggregated by categories and subcategories of explosive weapons.
For example, some sources indicate if harm to civilians occurred specifically in populated areas. Other sources may not make this distinction outright, but other relevant information can be used to determine independently whether or not recorded incidents occurred in populated areas. This might include determining whether there is a presence of civilians and civilian infrastructure by looking at population density and level of urbanisation, for example, or the presence of civilian infrastructure such as homes, schools, and hospitals.
The way that data systems define ‘explosive weapons’ similarly contributes to the ways their impacts on civilians will be interpreted, understood and reported. As such, collectively agreed understandings and definitions, and the promotion of substantial correlations in the use of terminology, can improve data effectiveness.
Definitions for these terms can be derived from technical and legalistic categorizations, definitions agreed among engaged communities of practice including civil society actors, or a combination of both.
Utilising technologies with the capacity to process large quantities of information
New opportunities to utilize the automation of data collection and processing through the application of artificial intelligence systems offers potential to speed up some elements of data collection and reduce the burden on human resources. Machine learning and natural language processing can offer the ability to scan and parse millions of online sources that include references to the impacts of explosive weapons.6
For example, the private technology company Dataminr, through its newly-launched AI for Good program, will work with Insecurity Insight, the UN Development Programme (UN) and other partners to provide AI technology to data collection efforts. For Insecurity Insight, this support will automate tasks and identify trends and patterns in data related to attacks targeting food systems, thus complementing Insecurity Insight’s existing work on aid security, health, and education.7
Finding broader contexts for harms defined in data systems
Standardized codes and methodologies for data collection allow for meaningful comparisons of data that help researchers, policymakers, and humanitarian organizations understand the scale of use of explosive weapons and severity of damage and harm in a consistent manner.
Impacts on civilians from explosive weapons use that are coded within data systems will nonetheless each remain context dependent. Many factors precede, and thus influence or determine, the outcome of the use of explosive weapons. The resulting civilian harm can be magnified by the concentration of population, for example, including as a result of displacement or 'besieged' or 'encircled' areas, by the age and deterioration of buildings and infrastructure, by the inability to find shelter, or the decrepit state of hospitals, services, or utilities.
A closer diagnosis of harms could better show the interaction between the direct impacts of explosive weapons, their longer-term effects and the preexisting contexts and conditions that contributed to those effects, to the extent possible. To that end, the United Nations Institute for Disarmament Research (UNIDIR) has developed menus of indicators that can be used and adapted to demonstrate quantifiable reverberating impacts of the use of explosive weapons.8
Sharing responsibility for proof with the users of explosive weapons
Prior to the Political Declaration, data collection was largely understood in the context of the work of civil society, the United Nations (UN) and the International Committee of the Red Cross (ICRC). Its purpose was to raise awareness of the humanitarian impact of explosive weapons use and to push for more effective protection of civilians.
The notion of data collection under the Declaration is expanded and makes clear the need for states and their armed forces to also understand the causes of civilian harm. This understanding can then be used to inform tactical changes and the development of policy to mitigate harm and to support accountability by identifying civilian harm incidents that require investigation.
Conclusion
Understanding both the challenges and possibilities of data collection on the direct and reverberating effects of explosive weapons use can support the efforts of a growing community of practice to collectively advance the understanding of the full scope of these impacts.
The Political Declaration provides a framework for this community of practice – including civil society, international organisations, and states – to work together towards this increased understanding, as its data collection provisions serve to set norms and establish good practice in recording not only deaths and injuries from explosive weapons use, but also the broader economic and social impacts.