• warning: file_get_contents() [function.file-get-contents]: php_network_getaddresses: getaddrinfo failed: Name or service not known in /var/www/web185/web/sites/all/themes/org/template.php on line 105.
  • warning: file_get_contents(http://urls.api.twitter.com/1/urls/count.json?url=http%3A%2F%2Foxfordresearchgroup.org.uk%2Fpublications%2Fbriefing_papers_and_reports%2Frecording_practice_policy_paper) [function.file-get-contents]: failed to open stream: php_network_getaddresses: getaddrinfo failed: Name or service not known in /var/www/web185/web/sites/all/themes/org/template.php on line 105.

Policy Paper: Towards the Recording of Every Casualty

  • Share
  • 0
Elizabeth Minor
22 October 2012

Everyone knows that in conflict people die, but they are not just numbers. They are people with dreams, with hopes, with families, with suffering, with all of that.

Quote from one the casualty recorders interviewed for our research


The 'Wall of Names' at the Kigali Memorial Centre, Rwanda, remembering Rwandan genocide victims. Source: Igor Roginek (2012) The 'Wall of Names' at the Kigali Memorial Centre, Rwanda, remembering Rwandan genocide victims. Source: Igor Roginek (2012)


This policy paper reveals the findings of the largest ever study of casualty recorders. The study, conducted by Oxford Research Group (ORG), demonstrates that the comprehensive recording of the deaths of individuals from armed conflict can and should be done.

The study is based on substantial research carried out our Documenting Existing Casualty Recording Practice Worldwide project, which is part of our our Every Casualty programme work. For two years, we conducted detailed research into the work of 40 organisations and individuals, who record the casualties of different conflicts across the globe.

This paper brings the results of this study to a wider audience. It makes recommendations to states, international organisations and civil society for the immediate improvement of casualty recording worldwide, and it shows the benefits and feasibility of casualty recording.


This policy paper can be downloaded as a PDF here, or read in full below.


Further Reading

From our research, ORG has also produced a collection of five practice papers, focused on technical issues in casualty recording practice. This collection is aimed primarily at current and future casualty recorders, to help them develop their work. It will also be of interest to others who wish to learn more about the process of recording and issues in recording practice:

Good Practice in Conflict Casualty Recording: Testimony, Detailed Analysis and Recommendations From a Study of 40 Casualty Recorders



Policy Recommendations and Analysis From a Study of 40 Casualty Recorders

Elizabeth Minor


Executive Summary

Many violent deaths from conflict around the world are either poorly recorded or not recorded at all. This policy paper argues that comprehensive recording of the deaths of individuals from armed conflict can and should be done. The paper demonstrates this from the results of detailed research by Oxford Research Group (ORG) into the work of forty organisations and individuals, who record the casualties of different conflicts across the globe. This research has also identified steps that can be taken now to improve the state of casualty recording worldwide (see the ‘Summary of Recommendations’). The research is the largest study of casualty recording practice ever carried out, to our knowledge. This paper sets out issues that must be taken into account in order to develop effective policy around casualty recording, according to the evidence produced by this research.

The main conclusions of this paper are:

Useful documentation of deaths from conflict can be done even during intense conflict and in repressive and dangerous environments

There is a range of ways in which casualty recording can be done under different circumstances (see section 2). The detail and certainty that a casualty recorder achieves in their records of deaths will depend on the context and the resources available to them, as well as their goals. A recorder’s work will be affected by: the types of sources of information that they can use and the investigations that it is possible to do; the intensity or stage of the conflict; and how much political space there is for recording deaths. These factors will mean that different approaches to recording need to be taken. However, all these approaches can produce robust and useful records (see section 2.2.2).

All types of casualty recording are valuable, and can support a number of different objectives

Casualty recording can support: the rights and recognition of victims and their families; fuller knowledge of the trends and consequences of conflict, which can help inform humanitarian response planning and violence reduction policies; and processes to uphold the law (see section 2.1.1). All approaches to recording will have uses (see section 2.2.2 for the different approaches to recording and their associated uses). The level of detail and certainty given by the data produced will vary between approaches. However, being able to achieve a narrower set of details about each case or a lower standard of proof does not mean that recording will serve no function. The highest standard of proof and detail is not always needed for every objective that recording supports (though the field of recording generally aspires to this standard). For example, our study has found that the baseline information about conflict incidents that some approaches produce was still of sufficient quality to be useful to other actors, such as humanitarian agencies for their assessment of a conflict situation. Approaches that give more detail about incidents and victims and to a higher standard of proof will support other purposes for which this level of sophistication is a fundamental requirement, such as legal accountability processes. 

Different approaches can connect: more detailed investigations can build on information collected earlier by other approaches

For comprehensive casualty recording, it is crucial that whatever recording is possible under the circumstances should be done. This is important for preventing information loss. Different approaches will also have uses that are necessary at different stages during and after conflict (see section 2.1.1). For example, immediately collecting and corroborating information available during conflict about deaths may sometimes provide only a limited record. However, this record can support humanitarian response planning by identifying areas of danger and need in close to real time. Collecting this information will also provide the basis for later, more detailed investigations and a more comprehensive record of individual victims. This more comprehensive record can contribute in the longer term to processes for upholding the rights of victims and their families.

There are steps that can be taken now to improve the effectiveness of casualty recording worldwide

States, NGOs, and other institutions and organisations that support or do casualty recording can take action now to improve work in this important field, and make recording more widespread and comprehensive. The recommendations of this paper are summarised on the next page and elaborated on in section 3. 


Summary of Recommendations

Drawn from and supported by our research into the work of casualty recorders, ORG makes the following recommendations for the immediate improvement of the state of casualty recording worldwide:

States should:

• Actively pursue the collection of all information about casualties when participating in conflict, and publish this information and share it with recorders as long as it is safe to do so;

• Raise their awareness of casualty recording, engage with its practitioners, and contribute to the development of the field;

• Where there are truth and reconciliation processes, integrate casualty recording and associated data into these.

Inter-governmental organisations and their agencies, both centrally and at country level, should:

• Share information about casualties with recorders, as long as it is safe to do so;

• Raise their awareness of casualty recording, engage with its practitioners and contribute to the development of the field.

All conflict parties should:

• Actively pursue and facilitate the collection of all information about casualties, and share this information with recorders as long as it is safe to do so;

• Not obstruct casualty recording, or those who collect information about casualties.

Global civil society should:

• Share information about casualties with recorders, as long as it is safe to do so;

• Raise their awareness of casualty recording, engage with its practitioners, and contribute to the development of the field.

Organisations that could use casualty information to benefit conflict-affected populations should:

• Make connections with casualty recorders, communicate data requirements, and use recorders’ information.

All casualty recorders should:

• Work together for joint standards for the field;

• Publish disaggregated information as long as it is safe to do so;

• Make connections with institutions that help realise recording’s benefits to conflict-affected populations.

These recommendations are steps that can be taken now, where they have not been taken already. Our research has shown that casualty recorders in civil society often operate where official efforts to record are seen as inadequate. These recorders address a need and show that recording can be done. Our research suggests that achieving the objective that every casualty is recorded will require both the development of current recording practice and global action by states and others at the highest level, to overcome the current obstacles and challenges to comprehensive recording.

This policy paper is based on information gathered by ORG about how the casualties of armed conflict are currently being recorded. The analysis and recommendations may also be relevant to a broader discussion of recording the casualties of all forms of armed violence1, but the focus of this study is armed conflict.


1. Introduction

1.1 The purpose and structure of this paper

Many violent deaths from conflict around the world are either poorly recorded or not recorded at all. This policy paper argues that comprehensive recording of the deaths of individuals from armed conflict can and should be done. The paper sets out important issues that must be taken into account in order to develop effective policy around casualty recording, according to evidence that ORG has gathered.

ORG has conducted a detailed study of the work of forty organisations and individuals that document deaths from armed conflict. Some also documented deaths from other forms of armed violence, but work on conflict was our focus. These casualty recorders worked to record deaths from conflicts across the world, in Europe, Asia, Africa and South America. Some worked in conflict zones; some worked from outside them; and some worked towards a full accounting of the deaths from conflicts where violence has now ceased. The Uppsala Conflict Data Program’s definition of armed conflict was used to identify the casualty recorders that should be included in the study.2

The study involved a survey of these casualty recorders that asked detailed questions about the recorder and their work. The questions covered areas such as: the definitions used by recorders in their work; their sources and confirmation methods; the challenges they faced and the things that helped them; how they released the information they collected; their aims and audiences; and how their work was used.

Almost all of the casualty recorders ORG surveyed were civil society organisations. Most were small, involving less than twenty people, and most recorded the casualties of intra-state conflicts.3 However our findings are relevant beyond civil society casualty recorders. This is because there are methodological issues and challenges in casualty recording that apply whoever the recorder may be. The range of benefits to recording is also the same. Additionally, the organisations included in this research that were not civil society-based casualty recorders were not found to be significantly different in their goals, methods or resources. Several of the civil society casualty recorders we interviewed did their work because of an absence of state action to adequately record casualties. The lessons learned from civil society based casualty recording are therefore directly applicable to states and inter-governmental organisations.

Using what ORG has learned about the practice of casualty recording from this study, this paper seeks to demonstrate why casualty recording is necessary and possible. Building on this discussion, the paper also makes recommendations for the immediate improvement of casualty recording worldwide. All the analysis and the recommendations given in this paper are conclusions drawn by ORG directly from the examples given and work done by the forty casualty recorders we surveyed. Readers should note that we promised full anonymity to all survey participants due to the dangerous environments in which some operated. Any details that could identify casualty recorders, including the country that they worked in, have therefore been removed from the examples and quotes from recorders given in this paper.

ORG is committed to the principle that every casualty of armed violence should be recorded (see ‘About Oxford Research Group’). We use the definition of armed violence given by the Organisation for Economic Co-operation and Development – Development Assistance Committee (DAC).4 Our focus is on deaths resulting from the use of weapons, in any situation where these go unrecorded. A major cause of unrecorded deaths from armed violence is armed conflict, and this study of recording practice focused on armed conflict as one specific form of armed violence. However, the analysis and recommendations given in this paper may also be relevant to a broader discussion of recording the casualties of armed violence.

After defining what we mean by casualty recording, this paper:

1. Discusses why casualty recording is important, by describing the principles that inform the work of casualty recorders, and uses of their work;

2. Describes how casualty recording can be done in different circumstances; shows why all approaches to casualty recording are important by discussing their different uses; and demonstrates how different types of casualty recording feed into each other towards a more comprehensive record. This section also proposes key standards for effective casualty recording that can be applied globally and to all approaches;

3. Makes recommendations for how the state of casualty recording worldwide can be improved, including actions to support existing recorders. These recommendations can be implemented immediately and are made to a variety of actors. The recommendations are based on and include analysis of the challenges that casualty recorders faced in their work. They are also based on the actions that casualty recorders said could help them work more effectively.

1.2 What do we mean by casualty recording?

For this study ORG looked at the recording of deaths from armed conflict only, though the term casualty can also include people who are injured.5 Within these boundaries, casualty recording has the following fundamental characteristics:

Incident or individual level documentation

Casualty recording involves either:

• Documenting the deaths of individual people from conflict violence (e.g. listing individual victims and the circumstances of their deaths), or

• Documenting separate events or incidents in which deaths from conflict violence occurred (e.g. listing dates and places of separate incidents of violence and the numbers killed in each).

Many recorders will document both these kinds of information: the collective picture of all the deaths that occurred in a particular incident, and the details specific to individual deaths. 

Minimum pieces of information6 

In terms of the information collected, at a minimum casualty recording means documenting the following details about each incident of violence recorded:

• The date

• The location

• The number of people killed

• A description of the type of violence involved (which will generally relate to the weapons that were used)

Many casualty recorders can and do consistently record more information than this. There is also a difference between the minimum details that a recorder requires to document a case, and details that they will record where these are available. At its most comprehensive, casualty recording means building very detailed knowledge of victims (including both personal details and information about affiliations, such as combat status), perpetrators, and incidents for every case. A wide range of specific details will be systematically collected.

Thorough, consistent, aiming to be as comprehensive as possible

Casualty recording means collecting and confirming information about deaths thoroughly, and with a consistent methodology. It will mean keeping a record of the sources of information used for each case (though these will often remain confidential to protect the safety of individuals). 

Recording will aim to give a picture that is as comprehensive as possible of the deaths from the type of violence that is being documented. Some casualty recorders may only record casualties in one particular area of a country in conflict, for example if it is dangerous for them to work in areas where they are considered outsiders. Some may only record violence committed by one perpetrator or from one type of weapon. Others may only record civilian deaths, or concentrate on establishing and recording the fate of the missing or the identification of unknown victims. The goal for all is a record that is as thorough and complete as possible, given the constraints of the situation that any recorder is working in. Recording will give a continuous, case-by-case record of deaths across the time period that a recorder is concerned with. Casualty recording can be done either during or after a conflict.

The numbers produced by casualty recorders about how many people were killed in a conflict (or how many were killed across different areas/periods of time, or how many civilians were killed) will be from a simple count of all the individuals, or deaths from incidents, that have been recorded. Therefore, casualty recording is different to approaches that calculate total numbers of deaths through statistical estimation based on sampling.7 This is a different field that can have different uses and was not our subject of study. Our interest was in approaches to incident or individual level documentation of deaths.

Public acknowledgment

A key component of casualty recording is that the information produced about incidents or individuals is made public in some way. If there is a threat to safety, releasing casualty information or certain aspects of that information should be delayed (though information might be confidentially shared for specific purposes that benefit conflict-affected populations). There may be other reasons to delay release, such as to inform families or ensure information is confirmed. However, an assumption of casualty recording is that the information produced is in the public interest, important to the public good, or useful to release for other reasons.


2. Why Casualty Recording Can and Should Be Done

The first part of this section (2.1) aims to show why casualty recording is necessary. It reports how the work of casualty recorders is used to benefit different objectives, and what casualty recorders said about the principles behind their work. The second part of this section (2.2) uses the work of casualty recorders to show that recording is possible. It looks at how recording can be done in different circumstances; the specific benefits of different approaches; and how different types of recording can feed into each other to create a more comprehensive record of deaths. Finally, to complete the picture of how useful recording can be done, some key standards for effective casualty recording are discussed.

2.1 Why record: the principles behind casualty recording and the benefits it can have

All the casualty recorders that we interviewed aimed to provide information about conflict deaths that was missing from the public record: Recorders observed a need for important documentation that was not being fulfilled. One common objective in generating this new information was to counteract misinformation or ignorance about casualties. Additionally, a shared belief running through the motivations of all the casualty recorders we interviewed was that collecting and sharing this accurate knowledge about human losses from conflict can achieve positive change, either in national or international policy, or for victims or the community. This was borne out by what recorders reported about how their work was used:

2.1.1 How are casualty records used?

Casualty recorders did not just aim to uphold important principles, as described in section 2.1.2. Casualty recording benefits a number of different objectives through the detailed case-by-case information it gives. Those we surveyed reported that their work was used for the following purposes, within their own organisations and by others:

Humanitarian response or planning

Casualty records, when produced in close to real time, can make a useful contribution to the on-going assessment of a conflict environment by humanitarian responders. Some recorders working during conflict reported that the information they collected was used:

• By local communities for their own early warning activities: Recording gave them knowledge about conflict events in their area;

• By UN agencies, international NGOs and others for their assessment of the conflict situation, and to inform response planning through the indications given about areas of danger and need;

• To contribute directly to the monitoring of civilian protection in conflict by UN agencies.


Some of these recorders used mapping software to display their casualty data in a way that was dynamic and relevant to those seeking ‘what, where, when’ information about conflict incidents.8 Visualising casualty data dynamically on maps was reported by several recorders to be an important way to bring their work to various audiences.

Policy and trend analysis

The results of casualty recording can be used for policy evaluation and the systematic analysis of conflicts. Governments and international agencies, media organisations and NGOs, and researchers in universities and think tanks used the casualty data produced by the recorders we surveyed.

Recorders frequently analysed patterns and trends in deaths from conflict, to evaluate the impact of the strategies that governments were taking, or to monitor and evaluate policies that were designed to reduce violence or deaths. A small number of recorders reported a positive relationship with governments, to whom they made policy recommendations and who used their casualty data for assessments of the conflict situation.

Recorders also reported that their data were used by international organisations such as the WHO, UNDP, World Bank and EU for research and assessment of conflict dynamics. One recorder also contributed information to the UN Monitoring and Reporting Mechanism on Children and Armed Conflict, which collects information on grave violations of children’s rights, including killings, for action by the UN Security Council.

Official purposes of the state

Casualty recording by non-governmental groups in some cases contributed to official records or counts of the dead. Sometimes a casualty recorder’s work contributed to specific official purposes, for example the allocation of war benefits by the state to the families of people who died.

Legal accountability and justice

Casualty recorders working both during and after conflict reported that their work contributed to criminal investigations and prosecutions. Domestic, regional and international courts used information about casualties produced by recorders. By implication, casualty recording can be used to help monitor compliance with the law by parties to conflict.

Recorders have contributed information to courts about individuals and about patterns of harm or victimisation in conflict. Detailed casualty recording gives information that is relevant to prosecutions on victims, perpetrators, and incidents, including forensic analysis in some cases. Casualty recording can help show how violence affects particular communities, or give information that can contribute to an assessment of whether the use of force was proportionate in particular incidents. Recording does not provide a legal analysis, but a body of evidence. The casualty recorders we surveyed contributed information or expert analysis to prosecutions or investigations either on request or on their own initiative. 

Casualty recorders that did less detailed recording, or produced records with a lower standard of proof, also reported that their work was useful to courts. This was either to give contextual information to prosecutions about the nature of the conflict, or to indicate where investigation might be directed.

Some casualty recorders undertook legal casework within their organisation, for which they used the information they documented about casualties. Apart from criminal accountability for legal violations, information about casualties and incidents documented by recorders was also used to seek compensation for victims’ families.

Recorders also reported submitting the information they collected to Special Rapporteurs in the UN system. The information produced by some was also used by Truth and Reconciliation Commissions (or similar processes).

The rights of families and memorialisation

Detailed casualty recording contributes to ending families’ uncertainty about the fate of their loved ones. For casualty recorders who worked on the identification of unknown victims this was the core purpose of their work, and realised by every case that they resolved. 

Other recorders sought as one of their objectives to generate a comprehensive record of human losses that would give victims and families recognition. The work of non-state casualty recorders has contributed to official memorials. Casualty recorders themselves have also created their own memorials, in the forms of books or online portals, to give a dignified memory to those who have died in conflict.

Just as casualty recording contributes to the activities described above, the results of many of these processes can also feed back into stronger casualty records. Information produced or collected by humanitarian actors can include details about casualties; court decisions can give the most robust confirmation of certain details; truth and reconciliation processes can give a baseline for casualty recorders to work from. Casualty recording and these other processes can be mutually reinforcing.

2.1.2 The principles behind recording: Why record and for whom?

Casualty recorders expressed both practical and moral reasons why they documented deaths from conflict. Together, these form a general rationale for why casualty recording is important from various perspectives. The main reasons to record (corresponding with the uses above) were:

• To generate data and analysis that could be used for evidence based policymaking or on the ground action such as early warning or humanitarian response;

• To build and raise awareness of an accurate picture of deaths from conflict, through which changes in discourse, perceptions and policy can occur, and so violence can be reduced or its recurrence prevented;

• For the value of an accurate record to political transition, transitional justice and future conflict prevention;

• Because of the need or right to know about human losses from conflict for societies, communities and families, and the value of transparency when it comes to the consequences of conflict;

• To dignify, recognise and memorialise victims and their families’ losses, and give victim communities a voice: some recorders reported the validation that witnesses and families experienced from telling their story to the recorder, and having it re-told by them;

• To inform the creation of a historical record that a post-conflict society can use to address the past, and for young people and future generations to have access to a historical record centred on the victims of conflict;

• For the contribution recording can make to accountability, justice and upholding human rights: Recording can reveal possible abuses through the comprehensive picture of harm it gives, as well as fulfilling families’ right to know the fate of their loved ones.

These reasons to record are about generating important and useful knowledge, the prevention of future harm, and redress for past harm. Recording might be done for one specific purpose or because of a range of broad principles: There are multiple reasons why it is important. Several of the casualty recorders we interviewed reported that they did this work because of the absence of adequate state initiatives to record, despite the clear importance of knowing about conflict deaths from the moral, legal and practical perspectives listed above.

Recorders had various intended audiences for their work, as seen in the reasons to record. These audiences could be within the country of conflict, international, or in the other countries participating in the conflict. This depended on the conflict dynamics, and the relationship of the recorder to the country of conflict. Some recorders were documenting casualties in their own country, some monitoring the actions of their government abroad, others operating on universal principles such as a concern for human rights worldwide. Ultimately, casualty recording is important and relevant to many different fields, principles, and groups. 

2.2 How recording can be done: useful casualty recording under different circumstances

ORG has found that useful casualty recording can be done even under very difficult circumstances, such as during intense conflict and in repressive and dangerous environments. There is a range of ways in which casualty recording can be done depending on the conditions, and the goals of the recorder. A recorder’s work will be influenced by: 

• The types of sources of information that they can use and the investigations that it is possible to do; 

• The intensity or stage of the conflict; and 

• How much political space there is for recording deaths. 

Achieving a comprehensive record of every casualty to a high standard of proof will not often be possible immediately. However, the information that can be collected will still be highly useful to various actors. If a more comprehensive record with all its benefits is to be achieved, whatever recording is possible should always be done, and followed up by more detailed investigations later. These ideas are explained in this section.

2.2.1 Recording as a connected range of approaches: from the outbreak of violence to post conflict

This study has shown that there is range of approaches in the field of casualty recording and that these are connected. This means that:

• There are different ways to record casualties that are possible under different circumstances, including at different points during or after a conflict; 

• These will give different levels of certainty or confirmation; 

• They will also give different levels of detail about victims and the circumstances of their deaths;

• All these different types of recording have their uses, which will often be needed at different points in time;

• And, work done at different stages of conflict or through different methods can provide a starting point for or feed into other more detailed types of investigations, which may be done later.

Further to separate approaches having their own uses, as well as providing a starting point for further investigations, collecting and recording all available evidence at any given time will help prevent information loss. Witnesses or documents may become unavailable, or memories become vague, if documentation is not started as soon as possible. If a robust and comprehensive record is to be achieved, it is also important that casualty recording should continue for as long as necessary. Especially in the case of the identification of unknown victims and the search for the missing, the work of recording can continue for many years. More exact information on victims, perpetrators and the circumstances of deaths can take decades to emerge into the public domain. However long it takes to discover, this information will still be important to families, communities and legal processes.

Different levels of confirmation or certainty in recording

Within the range of approaches to recording there are different degrees of certainty about casualty information that can be achieved. This should not be considered as a simple distinction between records of deaths that are ‘verified’ and those that are not: There is a scale of confirmation or certainty.

Whether a recorder has robust procedures to weigh up different sources and cross-check information is the first crucial factor in the level of certainty that their records should be considered to give. More certainty is given by records based on corroborated information and the systematic evaluation of contradictions between different sources, grounded in good knowledge of the context.9

The amount of certainty provided by different casualty records will depend to a great extent on the source material that a recorder had access to. In terms of the certainty of information given by a source, there are two important factors: 

• The level of detail given and the closeness to events of the source; and 

• The standard of proof that the authors of the source required in order to produce the documents or statements that are then used by a recorder.

When similarly robust evaluation procedures are applied, a record of the death of an individual that is based on a death certificate (in a context where these can be considered trustworthy), the testimony of eyewitnesses and family members, and the judgment of a court, gives more certainty than a record of an incident that is based on the independent reports of two news agencies stating that three individuals died in a certain place on a certain date, for example. 

This does not mean that records offering less certainty will not be useful or should be considered ‘unconfirmed’. The highest standard of proof is not needed for every purpose that casualty recording can contribute to. For example: Continuous open-source document-based recording during conflict that contributes to humanitarian response planning by giving indications of conflict dynamics does not need the same level of certainty that records contributing expert testimony to legal cases do. The former can also be produced in close to real time, which is necessary for the purpose of response planning, whereas the latter cannot.

A diagram to demonstrate the range of casualty recording

The explanation on the following pages builds up, step by step, a diagram (Figure 1) that illustrates the range of casualty recording practice and how it is connected. The discussion describes the variables in the range, then different scenarios in recording, and how these can link up. The complete diagram illustrates the idea described above that there are different types of casualty recording possible under different circumstances, all of which can be valuable and can contribute to each other in a connected range of practice. 

The diagram simplifies the field of recording, but does reflect real approaches and connections. It shows known uses of casualty recording from the work of those surveyed for this project. The explanation below should be read by numbered step with reference to the numbered areas on the illustrations.

Figure 1A The variables in the spectrum

• At different stages during and after conflict, (1) different types of recording, which offer different levels of certainty, will be possible. (2)

• What recording is possible will depend on the context, including the types and quality of sources available and types of investigations that recorders can do. (3)

• These will produce different types of results (4),and so different uses or benefits. (5)


Figure 1B A scenario during conflict

• During intense periods of conflict, certain sources might be available. (6)

• A recorder may not be able to independently investigate the information given by these sources, but can aggregate and corroborate them, (7) and so produce a database of conflict incidents. (8)

• Combined for example with mapping technology, such information can be useful to humanitarian response planners and conflict-affected communities for risk and needs assessments. (9)

• Such a database can also provide analysis that is useful for formulating policies to reduce violence, examining the limitations of current policies, and for academic/ wider research purposes. (10)


Figure 1C When the context changes

• With a change in the context (for example violence has decreased or stopped, or a recorder’s capacity to investigate has improved) different sources and possibilities to investigate will become available. (11)

• Work that has already been done, to corroborate infor- mation and create a database of conflict incidents from the sources that were available, can provide a baseline or starting point for new investigations. For example, this previous work (along with tips-offs or rumours) can provide indications about where further investigations should be directed, and gives records that can be built on and added to. (12)


Figure 1D More detailed investigation

• These new investigations may be detailed, on the ground investigations, which use new sources, or seek more detailed information from existing sources, to build a more detailed and certain picture of the human losses from a conflict. (13) 

• The result of such investigations could be a more comprehensive, detailed database about conflict casualties, which might build on an existing database of conflict incidents. (14)

• Such records can contribute to more detailed academic and policy analysis. They can also contribute to procedures that require a greater standard of proof or level of detail, such as assigning compensation, or the evaluation of the conduct of participants to conflict. (15)


Figure 1E The search for missing people

• Post-conflict, previous detailed on the ground investiga- tion into conflict incidents can assist work to search for and determine the fate of missing people, investigate graves and correctly identify unknown victims. (16)

• This work is needed to end families’ uncertainty about the fate of their loved ones, and to return their relatives remains to them. (17)

• The identification of unknown victims also contributes back to a more comprehensive picture of the human losses from conflicts, which are not limited to the missing. (18)


Figure 1F Completing the diagram: memorialisation and legal processes

• Information from a comprehensive and detailed database of victims or from unknown victim identifications can contribute to memorialisation. (19)

• Such information can also form the basis for expert evidence that contributes to investigations and prosecutions for crimes committed during conflict. Casualty records will not provide legal analysis, but they will provide information either on individual cases or on patterns of harm that can be used by others to make legal determinations. The level of detail given and standard of proof achieved by these records will be important for legal purposes. (20)

2.2.2 Recording in different conditions: five approaches

This section describes in more detail how work is done at different points on the range of recording described in section 2.2.1. We have identified five main approaches to casualty recording from our study. The sources and confirmation methods of these approaches, the circumstances under which they can be used, and the uses of the information that they produce, are summarised below. This model of approaches aims to show how recording is implemented under different conditions for different purposes, and again to emphasise the value of all approaches, no matter the level of confirmation they give.

The five models of recording are:

1. Document-based recording

2. Document-based recording with on-the-ground corroboration

3. Recording using an on-the-ground network

4. Multiple source investigation

5. Unknown victim identification

Document-based recording

This is recording that uses documentary evidence produced by others as its only source. Evidence might be from NGOs and other civil society organisations, media, social media, state records and inter-governmental organisations, and accessed publicly or privately. Records are made through cross-checking and evaluating the reliability of different documents. The level of certainty given ultimately depends on the quality of documents available. Most using this model during conflict saw their results as an undercount or a baseline, due to the limitations of the sources available to them. For example, many official documents were not accessible or in the public domain; the coverage and quality of media and social media can vary in extent. A baseline is useful for showing patterns in violence over space and time.

This type of recording can be done from both inside and outside the country of conflict, during and post conflict. Many of the documents used will be freely available online: Useful information can be collected whilst avoiding danger to recorders. Where the flow of information is restricted, or the media and the bureaucratic structures of the state give ineffective coverage, this approach will be less useful. Recorders using a document-based approach often did so because it gave maximum coverage on minimal resources.

Casualty recorders using this approach during conflict generally produced records continuously, some in close to real-time (within 24hrs hours of an incident). The majority of documents used were produced in close to real-time (e.g. media reports). Records were therefore used for risk assessment, informing humanitarian response planning, analysis of trends in conflict, and assessing and developing violence reduction policies. Document-based casualty information was used by courts as contextualising material, and to assess the need to investigate possible crimes. It was also used by media organisations. For those using this model in post-conflict situations, the range of documents available was greater, for example with more official records becoming available. A more comprehensive record of casualties was achievable, and these recorders contributed to official counts of the dead.

Document-based recording with on-the-ground corroboration

This is recording that mainly uses documentary sources as above, but which gathers some extra information or corroboration from on-the-ground sources or investigators.

Recorders using this approach had field workers or on-the-ground contacts who gave them occasional original information, e.g. on unknown cases in inaccessible areas, and/or an occasional extra layer of checking or verification especially on ambiguous or hard to confirm cases. The coverage given by this model primarily depends on the extent of documentary sources but can give increased coverage or better confirmation of certain cases, depending on the extent or depth of the on-the-ground corroboration available. Added on-the-ground corroboration can be useful in circumstances where, for example, media access is restricted to certain areas but a recorder’s contacts can reach information sources.

Like the document-only model, most recorders using this approach during conflict recorded continuously, some in close to real-time. Records were used for informing humanitarian response planning and community early warning, trend analysis, and monitoring and evaluating the actions of conflict parties and the effectiveness of policies to reduce violence. Governments also used the information produced by some using this model to inform their policies. It was also used by media organisations.

Recording using an on-the-ground network

This is recording that relies on a network of on-the-ground sources or investigators as its main source of information. This information may be checked against or added to by documentary sources.

A network could consist of civil society organisations (religious, NGO) in close contact with witnesses, families and other sources. Or, it could be built from individual paid workers or volunteers recording in their own areas. Some recorders rely on the workers of various organisations communicating the information about deaths that they routinely collect in the course of their primary work. Information received from on-the-ground sources is corroborated or reconfirmed by the recorder. The extent of coverage this approach gives depends on the reach of the network. Some were very comprehensive, some only operated in certain areas.

Many using this model worked in contexts that were extremely repressive and dangerous for anyone known to be attempting to document casualties. This model can have advantages in such an environment: It can operate with a low profile, and provides a way of documenting casualties if other information flows are restricted, unreliable or do not give good coverage (e.g. media is limited, NGO activity is restricted, official documents are inaccessible). Recorders using this model were based both in and outside the country of conflict. Being based outside the country had advantages for data security. However, deep local connections and high levels of trust are always needed to construct a network and collect information.

Recorders using this approach may be able to produce and share information in close to real time. This depends on the structure of the recorder’s communications with the network, which may be slowed in some dangerous environments. Records produced by this approach were used by humanitarian organisations for their assessment of the conflict situation, in legal processes and for compensation, by recorders to make submissions to international courts and Special Rapporteurs in UN system, and by media organisations.

Multiple source investigation

This is recording that uses a very wide range of sources with consistent on-the-ground investigation of cases. The aim is a comprehensive and highly detailed record.

All possible documentary and other information are collected in this approach. Collecting information from family members and eyewitnesses is prioritised. This approach aims to establish a wide range of detailed facts about every casualty to a high standard of proof. Evaluation and corroboration of multiple sources and original investigations are the confirmation procedures. This type of recording prioritises accuracy over speed, and will often have longer-term uses.

Most of those interviewed who were operating this model, were working post-conflict, though many started collating information during conflict using other models as listed above. This provided a starting point for more in-depth work. This approach benefits from the availability and accessibility of a wide range of official and other documentary sources. In less heavily documented contexts, those operating this model might rely on having a high profile as an organisation to reach all those with information. One recorder using this approach operated in a country where the state was hostile to such documentation. Some of their work had to be clandestine. Others suffered threats and intimidation from state and other parties. However, all recorders using this approach engaged their governments on the results of their casualty recording, and achieved some kind of response or dialogue. Some political space for recording may be necessary for this model to operate.

Recorders using this model during conflict may release information on casualties as they record it. However, given the depth of investigation involved, this will likely be in terms of weeks after an incident rather than days or hours. Information generated by this approach was used for memorialisation, to contribute evidence on individual cases or regarding patterns of harm to domestic and international courts, to make submissions to processes in the UN system, and by governments to allocate war benefits.

Unknown victim identification

This type of recording is separated from the other approaches by the use of forensic techniques to confirm the identities of the dead. The record of the dead created will be made up of these identified victims only: those who were missing, or buried in mass or clandestine graves.

As part of the process of investigation leading to the identification of unknown victims, recorders operating this model will investigate various sources. They will work with eyewitnesses and family members to collect ante-mortem data, and with communities to investigate where and how incidents happened and where graves are. They will establish comprehensive lists of the missing and their characteristics. Confirmation of cases is through the identification of unknown victims by forensic techniques including DNA identification. Their remains are then returned to their loved ones. These recorders will also generate detailed evidence on causes of death.

All those using this model of recording were operating post-conflict (though one operated in a situation of widespread organised criminal violence). An official request or permission from the state was needed by recorders using this approach to carry out certain key functions, such as undertaking exhumations. The work could not be done in a way that these recorders would consider ethical or legal without this cooperation. This approach therefore depends directly on some agreement from state bodies in order to function. Despite such agreements, recorders using this model often reported hostility or obstruction to their work from states, and from other actors unhappy with the threat that recorders’ work posed to them.

The core goal of this model is to end families’ uncertainty about the fate of their loved ones. These recorders invariably contributed their records to legal processes. Their work also contributed to truth and reconciliation commissions, and to memorialisation.

The quality of data produced by casualty recording

It is a core argument of this paper that all recording is useful. It should be clear from the discussion so far that being able to achieve a lower level of certainty does not mean either that recording cannot be done or that it will serve no function. Our study has found that recorders produce data that is of a sufficient quality to be useful to other actors irrespective of the range of sources available or the environment that they work in. 

Different types of sources will vary in quality between and within different contexts. For example, the media may extensively report on casualties in one country but have poor access in another; some publications will often be more reliable than others; and publications that are unreliable when reporting on certain areas may report well on others. There are no global rules or requirements for the sources that should always be used in order for recording to be valid. However, there will always be some information available, whether through social media, local networks, or official agencies, which, given robust evaluation, can be made into casualty records with continuous coverage. At a minimum, these records will always usefully contribute to the analysis of a situation of conflict, for example by communities, humanitarian actors, and policy-makers.

2.2.3 Key standards for effective recording

Having examined what recording looks like under different conditions in sections 2.2.1 and 2.2.2, this section summarises some key standards in methodology that can enhance the quality and impact of recording across the whole field. This discussion completes the picture of how useful recording can be done. This section is also relevant to some of the recommendations made later for how the state of casualty recording worldwide could be improved.

The list below of the key features of an effective recording system was developed by ORG through analysing the work of those who took part in this study, and looking at the principles in methodology that they reported were useful to their work.10 These features are important both to the process of recording and for building trust and credibility with target audiences. The features are that a casualty recorder’s work:

1. Is, and is seen to be, impartial and reliable,

2. Has clear, transparent definitions and inclusion criteria,

3. Has a transparent methodology with robust, multiple-stage checking procedures,

4. Is connected to local communities,

5. Uses multiple sources,

6. Publishes disaggregated incident/individual level information,

7. Is open to correction, or the addition of new information.

Casualty recorders may postpone the implementation of some of these features for safety reasons. In order to protect sources or others, sometimes it will not be safe to publish comprehensive casualty information, or detailed information about methodologies. However, it may be safe to share these privately with some key users of casualty data such as humanitarian responders and courts. 

All the features explained in detail below are generally achievable, and recorders in this study have expressed their value to casualty recording work.

Impartial and reliable

By impartial we mean that in the process of collecting, categorising and assessing information a recorder has no bias. There will be no discrimination regarding the victims that are included, and the categories a recorder uses to describe victims or incidents will not reflect any political prejudice. Being, and being seen to be, objective, independent and professional is important to recorders for building credibility with others. It is important:

• In order to for their work to be accepted by their intended audiences as a trustworthy representation of the situation;

• In order to decrease the credibility of attacks or smear campaigns by hostile parties on the recorder and its work. This is important for maintaining authority with the public and other audiences;

• So that recorders can operate more freely and without the suspicion of the state or other parties to a conflict, who might be wary of recorders’ motives or agendas. Also, in order to build contacts to share information with those who might otherwise have been suspicious or hostile;

• In order to build trust with victims’ families and witnesses, as both an audience and a source of information.

For most recorders surveyed, it was important to be seen as impartial and reliable as an organisation, as well as in their casualty recording work specifically. For some, being affiliated to an academic or research institution helped establish this. Others depended on the reputation and status their organisation had built nationally or internationally, in relation to casualty recording or other work that they did. 

For some casualty recorders in the research, being seen as neutral as an organisation was not important, and they were very clear about having strong political motivations to their work. These were reflected in the analysis they produced on their data. However, professionalism and objectivity in producing data was still crucial, in order to be able to use their casualty records credibly for their goals. These recorders gave examples of academics, courts, media and government bodies using their work.

Clear, transparent definitions and inclusion criteria

If casualty recorders state their definitions and inclusion criteria publicly, this can increase others’ understanding of their work, and ability to use it. It can also help build credibility and acceptance by following principles of transparency and openness.

Without knowing who is counted in casualty records (e.g. are they just civilians; are only direct deaths from violence included or are other deaths listed), it is hard for others to know what the records represent. It will also be difficult for others to determine how the records might relate to other counts or estimates of human losses from conflict. 

This is the same for definitions. Different interpretations of the term ‘civilian’ can explain differences between lists of casualties produced by different organisations in the same context. However, without explanation, these differences can affect the credibility of casualty recorders. A lack of clear definitions can also affect the ability of different people or organisations to integrate casualty information into their own records, or use it for their own analyses and other work.

Transparent methodology with robust, multiple stage checking procedures

Transparency in methodology again allows others to draw their own conclusions about how casualty records should be understood, and to evaluate how useful they are. It can help build trust and neutralise attacks on credibility by allowing recorders to clearly show the standards that they have been using in their work.

Robust confirmation procedures are important to all recorders and to the value of the information that they produce. A multiple-stage procedure of checking information, involving different individuals, is an approach that many recorders use in order to reduce errors in their work. It is good practice in recording to develop written guidelines in the form of a codebook and procedure containing the recorder’s definitions and methodology. This should be kept updated, e.g. with decisions on difficult cases.

Recorders may not find it safe to openly share their methodology under certain circumstances. For example, publicly revealing the existence of an information network could put sources in danger. However, under such circumstances the methodology might still be shared, where safe, with certain key individuals or organisations that used a recorder’s information to benefit conflict-affected populations, in order to establish trust and credibility.

Connected to local communities

Having a connection to local communities is important to effective recording. This ranges from having a detailed knowledge of the local context, which helps a recorder to interpret information, to close collaboration with local people and organisations.

For many recorders, individuals and organisations on the ground were key sources of information. To collect information effectively from such sources, a strong connection to local communities is crucial. 

The importance of local partnerships with other organisations for source information was emphasised by many recorders. These partnerships were both with organisations that documented casualty information and those that worked with local communities in other capacities. These could provide either information or connections/access to families or witnesses. 

The importance of knowing local languages and cultural sensitivities, and awareness of the difficulties and traumas that might be experienced by families and witnesses in giving accounts, were also noted. 

Trust was frequently brought up as key for recorders in gathering information, and was built from making connections with local communities. This might be done through building a national reputation with outreach and public awareness activities, having local workers build up relationships and reputation, or building relationships with local figures of importance such as religious leaders.

Where a recorder counted local communities as one of their audiences, a well-established connection or relationship helped with feeding back information, especially to marginalised groups. A connection to the community was also a source of credibility for some recorders.

Use of multiple sources

Using a variety of independent sources to document the same case is a basic feature of recording. Recorders considered the use of multiple sources important in order to overcome possibilities of bias or inaccuracy should any one source be taken in isolation. Multiple sourcing gives recorders more robust information by enabling cross-checking or corroboration, and is something all recorders attempted. 

Some recorders reported that different types of sources gave different areas of coverage or were better at giving different types of information: No one source or organisation comprehensively covered the entire conflict. Therefore, a more complete picture could only be created through combining as many different types of information as possible. For this purpose, partnerships or information exchange with other organisations or institutions that had access to different areas or types of information were crucial, especially where the information was not generally published.

On the different kinds of information that different types of sources might contribute, recorders gave various examples: Family members might be better at giving personal information, such as a victim’s profession, than an exact cause of death; military press releases may be very accurate on the weapons used in an incident, but less trustworthy regarding the combat status of victims; hospital records will be accurate on the cause of death but cannot give information regarding perpetrators.


Figure 2

Figure 2, based on the work of the recorders surveyed and examples they gave, illustrates these simple but important points. In a fictional country X, there is a non-international armed conflict, with a non-state armed group in control of a portion of the territory. State records therefore cannot provide comprehensive information. However, a UN agency may operate in part of the area out of state control. Both conflict parties will also have information to contribute to recorders that may not be accessible to others. Collecting from these sources will be important for the recorder in order to build a comprehensive picture. 

X also has a mountainous/forested area in which the reach of state institutions is poor. However, here, local NGOs and religious institutions can contribute information from the people they work with. This will again increase the coverage and detail of casualty records. Where information can be exchanged with medical sources, these can give both increased geographical coverage, for example in the areas not under state control, and different types of information, to build a more detailed picture. Where the operational areas of all these different types of sources overlap, the information can be corroborated for a more trustworthy record. 

Publish disaggregated incident/individual level information

An important principle for many recorders was transparency in their work. For over half of those surveyed, this extended to transparency in the information that they released about casualties: These recorders shared, or intended to share, lists of incidents or individuals with others, as well as numbers. 

Some did not share this type of information due to a lack of capacity: For example, they did not have the technical knowledge or time to post information on a website. Or, they felt that including all the data in the reports they published would make these reports too unwieldy. Some did not publish because of safety or confidentiality.

Those who did publish disaggregated information often kept certain details private for safety reasons. The names of victims were not released if this could put families in danger, for example. Certain information such as the names of witnesses was never released.

Some of the reasons why recorders felt it was important to release disaggregated information were:

• For the recognition, public acknowledgment or memorialisation of victims;

• For the purpose of building a historical record, which must be public and open;

• To build trust with audiences by showing what their numbers were based on, and that they were committed to openness and transparency;

• To counter accusations of bias or smears on the recorder, by showing all the information and how it was researched as fully as possible;

• So that, in principle, it was possible for others to verify the information, compare it against different records, or evaluate the conclusions a casualty recorder came to. This extends to families, who will be able to confirm the name of their loved one in a list, but would find it hard to know if that victim was included in aggregated figures.

Disaggregated information is important to those who use the results of casualty recording. The usefulness of casualty recording to others will often be in the granular information about different incidents and individuals that it can give, showing cases across location and time. Releasing disaggregated information is also important:

• For those recording in the same context to compare information with each other. Such exchange can help to build a more robust and complete record;

• In order for others to use the information for further recording or for the integration with other lists, as per the range of connected approaches to recording described in section 2.1.1.

Open to correction

Openness to correction has two components:

• As part of a robust recording methodology, new information should always be integrated into previously documented cases, so that the record is as accurate as possible. Some recorders had systems to actively seek corrections or additional information from others. For example, some invited people with information or queries to contact them via their website, or had other public outreach activities.

• Openness to correction and about the limitations of a recorders’ data is also part of the transparency and openness principles described above.

Several recorders remarked that the recording they did could never be finalised: New information and details would always emerge. To acknowledge this and make corrections was reported as a strength, not a weakness.

This paper has looked so far at why casualty recording is important and achievable, based on what ORG has learned from the work of recorders. It has examined the processes that casualty recording can contribute to, the principles behind it, and how recording can be usefully done. The final section of this paper, using information from our study, proposes recommendations for improving the state of casualty recording worldwide.


3. Recommendations to Improve Casualty Recording Worldwide

This policy paper’s recommendations are for improving the effectiveness and reach of casualty recording worldwide. They are steps that can be taken immediately. Our research suggests that achieving the objective that every casualty is recorded will require the development of current practice among casualty recorders, individually and collectively, and collaborative global action by states and others at the highest level.

This paper’s recommendations draw on the understanding of casualty recording practice gained from ORG’s research. The recommendations are based on the challenges that the casualty recorders we surveyed faced in their work. They are also based on the actions that casualty recorders reported could improve their situation, or improve recording globally. Each recommendation is accompanied by analysis or examples from the information collected in the research in order to demonstrate why the recommendation is made. The recommendations are made to different actors. This gives the structure of this section.

This policy paper is based on information gathered about how the casualties of armed conflict are currently being recorded. The analysis below may also be relevant to a broader discussion of recording the casualties of armed violence,11 but the focus of this study is in armed conflict. 

Following a general note on the need for information sharing, ORG makes twelve specific recommendations for the immediate improvement of casualty recording by states, civil society and international organisations worldwide.

A general note on the need for information sharing partnerships

The recommendation to share information about casualties with recorders is made to several different actors. This sharing is necessary in order to produce the most comprehensive, robust and useful record:

Many different types of organisations and institutions collect information that is useful to casualty recording. Health, humanitarian, and civil society organisations, state and inter-governmental institutions all encounter or record information about deaths. This information can be and is often used by casualty recorders to create a centralised record. As described in the previous section (2.2.3), combining information from multiple sources that operate in different areas increases the coverage and quality of the information produced by casualty recording, and minimises bias. Information sharing partnerships are needed for maximum coverage, as no one organisation can generally cover an entire conflict area. Where various different organisations and agencies are operating to collect similar information, using information already collected by others can also avoid the problem of witnesses becoming frustrated with repeatedly giving the same information. This can be a challenge for casualty recorders.

How to share information safely between institutions – whether this can be done publicly, privately, or at all without causing risk to the institutions, their information sources or their services – will vary according to context. Guidelines or standards will be needed if this practice is to be universal. However, ORG’s research discovered precedents for all types of institutions sharing information with casualty recorders formally or informally – central and local state institutions, the agencies of inter-governmental organisations, and civil society organisations. Any information given to recorders should always be shared at the level of incidents and individuals. This level of detail is necessary for casualty records to be established12 and for information to be corroborated against other sources.

3.1 For states

3.1.1 When participating in conflict, states and their local agencies should actively pursue the collection of all information about casualties, and share this information publicly and with recorders as long as it is safe to do so

States participating in conflict will often have access to information that few or no other organisations will. For example, armed forces will sometimes be the only witnesses to events. Because of this, if a comprehensive and robust record of casualties is to be achieved, all parts of the state from local to central institutions should actively pursue the collection of information about casualties. We recommend that this information is routinely shared with casualty recorders and published, as long as it is safe to do so. This information should be shared at the level of incidents and individuals.

Several casualty recorders reported that greater transparency from governments over casualty data would both be highly beneficial to their work and was important in principle. Data should also be shared across borders where conflicts were or became international: Information about casualties should be available to all whom it is relevant to. This includes those in other states who are recording the deaths of citizens of their country abroad. The data shared may not be complete and information from any source always requires further investigation and corroboration. However, all sources of information about casualties must be integrated if a comprehensive record is to be achieved.

State practice

Some casualty recorders that we surveyed experienced positive information relationships with the central governments of the country of conflict. Post conflict, some recorders reported the sharing of full databases by the military, missing persons agencies, and reparations agencies of the state. Others, both during and after conflict, reported the sharing of information by ministries for war veterans, defence, the interior, foreign affairs, missing persons, and the ombudsman. Some recorders had specific formal relationships with central government in order to do their work, usually when forensic work and exhumations were involved. Some states also had dedicated ad hoc or on-going projects to record casualties, with either published results or information shares with other recorders. One recorder working during conflict reported the routine exchange of information and central casualty databases with the military and police. This came about after the recorder had been publishing their results for several years with high public interest. The government chose to respond to this by publishing their own statistics from their own investigations, based on the standards of the recorder’s database. 

In contexts where these positive relationships existed, recorders noted a connection to the exercise of the rule of law. In some cases the governments concerned were interested in developing progressive policies for violence reduction. These examples show that the sharing of information by states with recorders is possible, where the public interest of recording is accepted. Official information-sharing agreements, recognising the legitimacy or status of casualty recorders, and giving them access to investigate are important steps that states could take.

The data shared by central government was often reported by recorders to be useful, though some also reported that it was affected by political agendas. As a result it was sometimes too biased to use. Some reported trying to access information about casualties through Freedom of Information procedures, but found that there was official suspicion of recorders asking for information, resulting in delays, slow or partial information release, or refusal.

Some recorders noted that certain types of information were more likely to be released by governments than others. One reported that there was very good road accident data publicly available, suggesting bureaucratic capacity, but very little data on violence. Others noted that the information governments published on casualties often concerned deaths caused by other parties to the conflict, rather than casualties that they caused themselves. Where governments had signed up to international obligations that included monitoring casualties, for example on landmines, structures were more likely to be put in place to do this monitoring. It may, however, be more politically straightforward to monitor casualties from specific causes, such as banned weapons that the government is not implicated in the use of, than to make a broader commitment to monitoring all types of casualties.

Most casualty recorders that we surveyed reported secrecy from states, a lack of recording, propagandistic statements about casualties (seen as reporting an interpretation favourable to the state, rather than the facts about deaths), or having no way of finding out if the government was recording or not. Sometimes, it was possible under these circumstances to discover that casualty data was being recorded when it became available through unofficial leaks. The general observation in these contexts was that the state did not see the release of data as in its interest. There were fears that soldiers might be discouraged from military action, that compensation cases would be generated, or that the state’s image would be harmed in other ways. There may also be concern regarding data protection and laws on the release of archives. If casualty recording is to become universal, these issues must be addressed. It has to be recognised that states are not impartial observers to conflict in many circumstances, and will sometimes have a conflict of interests between the benefits of releasing and the benefits of withholding or distorting casualty information. However, some recorders noted that by withholding the information they had, states damaged their image and credibility, and opened up a space for propaganda that others could then fill.13

The responsibility and capacity of the state to record

Whether or not they have the will to collect and publicly release information about casualties, some states do not have the capacity to do this. It was reported in some contexts that states did not have the bureaucratic capacity or reach, especially in areas of conflict, to collect and share information.

Some recorders reported that they did their work because the state appeared unable or unwilling to record, or at least to do so publicly. ORG proposes that casualty recording should be an internationally accepted obligation, and that the ultimate responsibility to ensure casualty recording is done lies with states. The principle that the state should take responsibility for recording was expressed by several of the recorders we interviewed. However, as described in the previous section (2.2.3), impartiality and trust is key to the most effective and beneficial recording. This means that the state should not necessarily be the only body making casualty records. The state should take responsibility for ensuring that recording happens, as well as investigating casualties to the extent possible and releasing all the information it has. However, for the sake of trust in the records produced, or because of limited capacity, recording should sometimes either be entrusted to another institution or organisation, or done through collaboration between the state and others. Whatever arrangements are appropriate in the context, cooperation between the state and other actors will always be essential to the most effective and comprehensive casualty recording.

Local cooperation

Cooperation with state agencies was often reported by recorders to be more successful at a local level, especially when engaging institutions that collected information about casualties during routine bureaucratic procedures. Useful information was shared with casualty recorders, formally or informally, by:

• Police

• Military and intelligence agencies

• Courts

• Local government officials

• Health and medico-legal institutions (which may also be separate from the state)

Recorders most frequently reported sharing from medical organisations or their personnel. In one case, a health institution collecting casualty information for their emergency response planning shared this, including publishing numbers, which was useful to other recorders. For some agencies, there may be both little effort and little risk in sharing information that they are routinely collecting in the course of their work anyway.

3.1.2 States should raise their awareness of casualty recording, engage with its practitioners, and contribute to the development of the field 

States who recognise the value of casualty recording should raise their awareness of this work and support the activities of the field, in addition to collecting and releasing information.

Supporting threatened practitioners

Several of the casualty recorders that we interviewed experienced threats from conflict parties and their supporters because of their work. They reported that the support of representatives of other countries (e.g. locally-based diplomats) was important for countering this intimidation. This was most useful when expressed publicly, for example in press conferences. Expressing this support showed that the recorder had high profile supporters. This discouraged those intimidating the recorder from acting on their threats. States who recognise the value of recording should publicly support threatened recorders and the work of the field around the world.

Funding casualty recorders

Recorders reported that a lack of funding for their activities was a serious issue. Some reported that because the international donor community did not see their country as a priority, they could not access funds. However, casualty recording is a long-term commitment and is important in every context. States who support casualty recording should prioritise its funding globally. States should also fund casualty recorders in their own countries, especially where they do not have the capacity to do this work themselves. However, in some contexts the funding of casualty recorders by any state would not be appropriate, as it would compromise the perceived impartiality of the recorder. 

Integrating recording into international action

States that recognise the value of recording are encouraged to raise this issue in international forums. One recorder noted that international initiatives, including UN Security Council Resolutions, are often passed with a requirement for the monitoring of violence, but with no resources or structures put in place to ensure that this happens. Recording casualties is a vital part of monitoring violence. States who support casualty recording should advocate for these calls to be properly structured and resourced.

3.1.3 Where there are truth and reconciliation processes, integrate casualty recording and associated data into these

Many recorders saw casualty recording as a crucial step in post-conflict truth, reconciliation and transitional justice processes. Where truth and reconciliation commissions (or similar bodies) are initiated, casualty recording has in the past and should in the future contribute to these. One recorder noted that casualty recording and the work of truth and reconciliation commissions could interconnect better towards a comprehensive record of deaths. For example, if processes are set up whereby reduced sentences are given in exchange for testifying at the commission, information about casualties (including, for example, the location of graves) could be made a central part of that exchange.

3.2 For inter-governmental organisations and their agencies

3.2.1 Share information about casualties with recorders, as long as it is safe to do so

As with others who collect or encounter information about casualties in the course of their work, inter-governmental organisations and their agencies are encouraged to share this with recorders, at the level of incidents or individuals.

Safe sharing and humanitarian space

Sharing information safely between institutions without causing risk or damage to the work of the informant will require guidelines or standards if it is to be globally implemented. Humanitarian aid agencies, which are often the agencies of inter-governmental organisations and so are discussed here, have specific concerns around information sharing that were reported by casualty recorders we surveyed. These concerns were that sharing information with casualty recorders, who might then release it, could endanger humanitarian space. Conflict parties and other hostile forces might no longer see the humanitarian agencies as neutral, and so would not grant them access for their humanitarian operations. Where the information could be traced back to the humanitarian agency, and where conflict parties have clear hostility to recording, this is a serious and substantial concern. Casualty recording should never further endanger life, and information sharing should not be done where there are threats to safety. Recorders must find alternative sources, or postpone their activity until the security threat is reduced.

However, in many circumstances casualty recording will not endanger humanitarian agencies and their services. Some recorders reported successful information exchanges with humanitarian agencies during conflict. These relationships were mutually beneficial. The casualty recorder would provide information that the agency would use to contribute, for example, to security briefings, assessing civilian protection, and early warning. The agency would in turn give information to the casualty recorder. The key in these information partnerships was to establish trust and legitimacy between the casualty recorder and the agency at the local level.

The practice of inter-governmental organisations

There was a significant variation reported by casualty recorders in the sharing of information by inter-governmental organisations and their agencies. Sometimes public lists or reports about casualties or the missing were issued by organisations, such as the ICRC, UNOCHA and different UN country missions. This is a very useful practice that should be encouraged, and the public profile of these activities should be raised. In other places, there was a refusal to share any information. There was not necessarily a pattern here. In one case, one organisation worked very closely with a casualty recorder in a post-conflict setting, performing quarterly information shares and publishing information on a website. In another post-conflict country, the same organisation did not cooperate at all with the recorder we interviewed, although it was a well-established organisation in its country.

The impression we received from recorders was that central policy in inter-governmental organisations was generally against information sharing, but local officials made their own decisions based on what they thought was beneficial in the context. It could be useful if discussions on sharing information about casualties were held centrally within these international organisations and their agencies. In general, recorders called for sharing and publishing from these organisations.

3.2.2 Inter-governmental organisations should raise their awareness of casualty recording, engage with its practitioners and contribute to the development of the field

Inter-governmental organisations and their agencies that recognise the value of casualty recording should raise their awareness about this work. Where the results of casualty recording could be institutionally beneficial (for example, for informing humanitarian response planning), organisations and their agencies should engage with casualty recorders (see recommendation 3.5 below).

As mentioned in recommendation 3.1.2, international action that contains a requirement for monitoring violence, including that authorised by the Security Council, should also contain provisions for the proper implementation of this monitoring. Recording casualties is a vital part of monitoring violence. Inter-governmental organisations that recognise the value of recording should advocate for this monitoring to be properly resourced and structured.

3.3 For all conflict parties

3.3.1 Actively pursue and facilitate the collection of all information about casualties, and share this information with recorders as long as it is safe to do so

In the course of their actions, state and non-state parties to conflict should actively pursue the collection of information about casualties. This information should be shared with recorders at the level of incidents or individuals, as long as it is safe to do so. 

The capacity of newly established non-state armed forces to record will be low in some cases, as casualty recorders noted. However, non-state armed groups should not be excluded from expectations of casualty recording. Several recorders reported that non-state armed groups recorded and published the casualties of their own forces online, in statements to the press, on posters, or through murals. These releases were a useful source to recorders, though they sometimes contained misinformation from the armed group, aimed at improving their public image. A small number of recorders also reported that they were able to liaise with non-state conflict parties for information about civilian deaths in incidents in which they were involved. Again, this was a very important if not always reliable source of information, as with similar exchanges with state parties to the conflict. 

3.3.2 Do not obstruct casualty recording, or those who collect information about casualties

Several of the recorders we interviewed reported that restrictions of access to certain areas by conflict parties were a challenge in their work. These restrictions were placed on the media, or on any outside investigator including recorders themselves. Both state and non-state conflict parties should allow and facilitate the access of casualty recorders and other investigators to areas of conflict. One recorder reported making a formal agreement with a non-state conflict party to recognise their documentation activities in the conflict area.  

3.4 For global civil society

3.4.1 Share information about casualties with recorders, as long as it is safe to do so

Civil society organisations should share any information they have about casualties with recorders, as long as it is safe to do so. Information should be shared at the level of incidents or individuals. Recorders reported that working through local information partnerships with civil society organisations that did not have a core human rights or documentation focus sometimes meant that recording could be done without attracting undue attention from conflict parties. Such activity must follow a careful assessment of whether it could damage the partner organisation’s core work or be detrimental to their safety.

Where the work of casualty recorders would benefit the work of civil society organisations with conflict-affected populations, connections should be made between these organisations and recorders (see recommendation 3.5 below).

3.4.2 Raise their awareness of casualty recording, engage with its practitioners and contribute to the development of the field

Civil society organisations worldwide that recognise the value of casualty recording should take steps to support the field. These organisations should raise their awareness about this work, support initiatives on this issue and engage with casualty recorders. 

Local civil society organisations frequently help recorders to connect with communities, both for the collection of information and in order to feed their work back to those affected by conflict.

Where casualty recording practitioners are threatened, public support from other civil society organisations, especially large international NGOs, was reported to be important in discouraging intimidation.

Local and international civil society organisations supported some of the casualty recorders we surveyed financially, with communications equipment or with software or other technical expertise. Civil society organisations that support casualty recording and have the capacity to do so should consider sharing resources and expertise with recorders.

3.5 For organisations that could use casualty information to benefit conflict-affected populations

3.5.1 Make connections with casualty recorders, communicate data requirements and use recorders’ information

The recommendations of this paper encourage transparency and public access to information about casualties. For casualty recorders, the core purpose of transparency and public access is to benefit conflict-affected populations, directly or indirectly. We recommend that the actors below should engage with casualty recorders and use their information for this purpose.

This engagement will include these actors specifying what data is needed from recorders. Different institutions will be able to use casualty data offering different standards of proof and detail (see section 2 for a discussion of the different approaches to recording and their associated uses). Some institutions may require that data has gone through certain processes of confirmation in order to integrate it into their work. Data requirements should be defined and agreed so that recorders can show they conform to these standards. Procedures for the safe sharing of data should also be developed.

Institutional links and relationships of trust should be built between recorders and the organisations that can use casualty data to benefit conflict-affected populations. The trust that should be built is institutional, and in the capabilities and methodologies of the casualty recorder and their data. It can be developed initially through a recorder sharing their methodology and definitions, and demonstrating the types of sources that they use (without threatening safety or confidentiality).

As set out in section 2.1.1 on the uses of casualty recording, the following actors can use information about casualties for the benefit of conflict-affected communities, and should engage with recorders over their data:

• Humanitarian actors

• Inter-governmental organisations, including agencies concerned with human rights

• Domestic, regional and international courts

• Local and international civil society organisations

• Researchers, research institutions and media organisations

• Policy-makers

• State bodies that assist victim communities, for example in the allocation of war benefits

3.6 For all casualty recorders

Those who record casualties may be NGOs or civil society organisations (including media, academic and research organisations), civil society collectives, inter-governmental organisations or state institutions. These recommendations are for all types recorders across the field.

3.6.1 Work together for joint standards for the field

Several recorders noted that the field of casualty recording could be strengthened and given greater legitimacy and recognition through developing common standards. As set out in section 2.2.3, ORG’s study has identified certain key standards that can contribute to a robust methodology and legitimacy with audiences. All recorders can adopt these. As also discussed in this paper, there are several different but equally legitimate approaches to casualty recording (see section 2.2). Any standards developed must be specific and appropriate to the methods that a recorder is able to use in the context.

ORG coordinates a network of civil society based casualty recorders, the International Practitioner Network (IPN).14 One purpose of the network is to discuss and develop common standards. The IPN’s discussions of this topic will build on the results of ORG’s study of casualty recording practice.15 Discussion on standards will also require the direct expertise of casualty recorders and others, such as the users of casualty data. Casualty recorders and other experts are invited to consult with ORG on this process.

3.6.2 Publish disaggregated information as long as it is safe to do so, and make connections with institutions that help realise recording’s benefits for conflict-affected populations

As mentioned in the discussion of the key standards for effective recording (section 2.2.3), the release of incident or individual level information about casualties is of crucial importance as long as this does not endanger sources, families, or others. Given the various uses of casualty recording that discussed in section 2.1.1, it is equally crucial that the information produced by casualty recorders is connected with the institutions and organisations that can help realise the full benefits of recording for conflict-affected populations. It is also important that recording is fed back to affected communities in a way that is beneficial. As one recorder pointed out, it could be dangerous to discover that your neighbour was possibly implicated in war crimes, where this was not coupled with other processes.

Ensuring that casualty recording is recognised as useful, and that it is connected to all the positive processes that it benefits, is a challenge for the field. We found in our study that many recorders concentrated on producing data without ensuring that their work had an impact. This was often out of necessity because of resource constraints. However, given that casualty recording is important and relevant to many different fields, principles, and groups, this must be overcome.


Final Thought

This report has looked at why casualty recording is necessary and how it can be done, using evidence from casualty recorders. Many of the recorders ORG surveyed did their work because action by the state and others was seen as insufficient. The current global condition of casualty recording, therefore, has to be considered inadequate. This report has made recommendations for strengthening the state of casualty recording worldwide right now. As several casualty recorders observed, however, long-term collaborative global action between all relevant actors at a national and international level will be necessary to ensure that every casualty is systematically recorded.



1 See section 1, p3 for definitions armed conflict and armed violence

2 “An armed conflict is a contested incompatibility that concerns government and/or territory where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle-related deaths in one calendar year.” Uppsala Conflict Data Program, (Date of retrieval 16 September 2012), www.pcr.uu.se/research/ucdp/definitions/, Uppsala University Department of Peace and Conflict Research

3 As well as this policy paper, ORG has produced the collection ‘Good Practice in Conflict Casualty Recording: Testimony, Detailed Analysis and Recommendations from a Study of 40 Casualty Recorders’ from the results of the survey. The collection gives a detailed treatment and reflection on different themes in recording practice. It is aimed at casualty recorders and others interested in these issues. See www.oxfordresearchgroup.org.uk/publications/briefing_papers_and_reports/casualty_recording_practice_collection. For a detailed description of how the survey was conducted, see the collection’s ‘Appendix on Survey Methodology’. For a more in depth treatment of the types of casualty recorders surveyed, see the paper ‘An Overview of the Field’.

4 Which states that armed violence is “the use or threatened use of weapons to inflict injury, death, or psychosocial harm”. www.poa-iss.org/kit/2009_OECD-DAC_Guidlines.pdf OECD, 2009, p.28 (Date of retrieval 16 September 2012)

5 Most of those surveyed recorded injuries. Injuries are a major part of the overall burden of armed violence, and in some cases lead to death. It is good practice when recording deaths to follow up cases of injury.

6 Oxford Research Group would like to thank participants at the conference ‘Recording Every Casualty’ in September 2011 for the inspiration their discussions gave to this section (see www.oxfordresearchgroup.org.uk/news/2011/10/first_conference_internation...).

7 For a discussion of this type of work, see for example Michael Spagat, ‘Estimating the Human Costs of War: The Sample Survey Approach’, www.socsci.uci.edu/~mrgarfin/OUP/papers/Spagat.pdf, (Date of retrieval 16 September 2012), and forthcoming in Seybolt, T., Aronson, J., & Fischhoff, B. (eds.) Counting civilian casualties: An introduction to recording and estimating non-military deaths in conflict (Oxford: Oxford University Press, in press).

8 Casualty recorders across the field used mapping software to display geo-located information, using either custom-built platforms, visualisations developed from Google Earth or tools such as Ushahidi.

9 For a detailed discussion of the issues and nuances involved in evaluating sources, see the paper ‘Evaluating sources in casualty recording’ in the collection ‘Good Practice in Conflict Casualty Recording’ www.oxfordresearchgroup.org.uk/publications/briefing_papers_and_reports/.... Some key features of a robust casualty recording methodology are set out in section 2.2.3 (page 16) below.

10 This list also builds on conclusions drawn by a previous working paper by Oxford Research Group. See Jacob Beswick, ‘Working Paper: The Drone Wars and Pakistan’s Conflict Casualties, 2010’, 2011, http://www.oxfordresearchgroup.org.uk/publications/briefing_papers_and_repor...

11 See section 1, p3 for definitions armed conflict and armed violence. 

12 See the definition of casualty recording, section 1.2 p4 above.

13 ORG, with a senior officer of the British military, has previously analysed the topic of the potential advantages to militaries of recording casualties. See Hamit Dardagan, John Sloboda and Richard Iron, ‘In Everyone’s Interest: Recording All The Dead, Not Just Our Own’, British Army Review No 149, 2010, www.oxfordresearchgroup.org.uk/publications/reports/everyone%E2%80%99s_i...

14 See www.everycasualty.org/practitioners/ipn

15 The collection ‘Good Practice in Conflict Casualty Recording’, www.oxfordresearchgroup.org.uk/publications/briefing_papers_and_reports/casualty_recording_practice_collection produced from this study contains analysis and recommendations that are intended to assist recorders in developing their work. Casualty recorders may find it useful to refer to the collection.



The 'Wall of Names' at the Kigali Memorial Centre, Rwanda, remembering victims of the Rwandan genocide (© Igor Roginek)

Women at the Srebrenica-Potocari Memorial and Cemetery for the Victims of the 1995 Genocide, Bosnia and Herzegovina. (© Adam Jones, bit.ly/SM6Wz8)

Relatives with identified remains of their families in Putis, Ayacucho, Peru (© Equipo Peruano de Antropología Forense)

An exhibition of photos commemorating casualties among rebel forces in Misrata, Libya, 2011. (© Richard Moyes/Article 36)

Families of people killed in the conflict in Kosovo providing information about their loved ones, at a public presentation by the Humanitarian Law Center in Prizren. Such contributions are invited in order to make a record that is as comprehensive and accurate as possible. (© Humanitarian Law Center)

The website of LRA Crisis Tracker, a collaboration of Invisible Children and Resolve. The Tracker records casualties caused by the Lord’s Resistance Army (LRA), which are displayed on this dynamic map. Casualty and other data can also be downloaded from the website.

The launch of the first volume of the Kosovo Memory Book, a joint effort of the Humanitarian Law Center in Serbia and the Humanitarian Law Center Kosovo. The book documents and memorialises people who died in the conflict in Kosovo. The call of the project is “let people remember people”. (© Humanitarian Law Center)

The International Commission on Missing Persons morgue in Tuzla, Bosnia and Herzegovina. (© Laila Zulkaphil/The Advocacy Project bit.ly/QbXjpm)

A book documenting the clothes worn by unidentified victims found in clandestine graves, at the International Commission on Missing Persons in Bosnia and Herzegovina. (© Laila Zulkaphil/The Advocacy Project, bit.ly/UzDmsN)

Before publishing the Kosovo Memory Book (see p5), public presenta- tions were made around Kosovo. Communities were invited to learn about the project, check whether loved ones they had lost were in the database, and make sure that details about them were correct. (© Humanitarian Law Center)

Mural commemorating casualties among rebel forces in Misrata, Libya, 2011. (© Richard Moyes/Article 36)

The form used by Documenta, Croatia, to record information about casualties. (© Igor Roginek/Documenta)

Fieldwork to document human losses in Croatia (© Igor Roginek/Documenta)


About the Author

Elizabeth Minor is the Research Officer of the Every Casualty programme at Oxford Research Group. She was the principal researcher on the two-year research project (2010-12) from which this paper and the collection ‘Good Practice in Conflict Casualty Recording: Testimony, Detailed Analysis and Recommendations from a Study of 40 Casualty Recorders’ were produced. The project was directed by Hamit Dardagan and John Sloboda (Co-Directors of the Every Casualty programme at Oxford Research Group) in collaboration with Michael Spagat (Royal Holloway University of London) and Madelyn Hsiao-Rei Hicks (Institute of Psychiatry, King’s College London).


Thanks to

All the reviewers of the research and its outputs, both internal and external to Oxford Research Group, who offered useful comments and support.

The forty casualty recorders who gave their time to participate in the survey, for the rich and valuable information they provided us about their work.


A Note on Quotes and Pictures

All the quotes given in this paper are from the casualty recorders we interviewed, used with their permission.

Some of the pictures we have used in this paper to illustrate casualty recording are from our partners, including members of the International Practitioner Network (IPN). This does not mean that these organisations or the IPN necessarily endorse this paper, whose content remains entirely the responsibility of Oxford Research Group. It also does not mean that the organisations that provided images used for this paper participated in this survey.



The Federal Department of Foreign Affairs of Switzerland and the United States Institute of Peace funded this study.

We welcome feedback and comments on our work. Please direct all enquiries about this paper or the study that it is a part of to [email protected] 


Design for the PDF

Tutaev Design


Date of Publication

October 2012

Copyright © Oxford Research Group 2012

Some rights reserved. This briefing is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Licence. For more information, please visit http://creativecommons.org/licenses/by-nc-nd/3.0/ 



Our publications are circulated free of charge for non-profit use, but please consider making a donation to ORG, if you are able to do so.

Copyright © Oxford Research Group 2016. Some rights reserved. This publication is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Licence. For more information please visit http://creativecommons.org/licenses/by-nc-nd/3.0/.