Introduction
Accountability is a key concept for humanitarian actors, but applying and fulfilling it is not without challenges. This article explores these challenges, the choices that need to be made and their consequences, as well as emerging ideas on how accountability can be met, by using the specific example of humanitarian protection activities in armed conflict settings.Footnote 1
As will be explored in greater detail below, accountability is currently understood as a largely technical exercise that explains to external stakeholders – i.e., donors and the affected population – how a humanitarian actor has used its resources and achieved its intended results. In the case of humanitarian protection activities – the example that this article is using – this means being accountable for ensuring that the rights of individuals affected by armed conflict are respected and that individuals are protected from the negative effects of war, which in turn often implies a change in the behaviour of parties to an armed conflict, or even larger social changes in the societies in which they live. With accountability focused on obtaining such broad change, this article argues that consequentialist ethical perspectives on accountability currently prevail. To quote Thomas G. Weiss, “consequentialist ethics are essential”.Footnote 2 This leads to the key question posed by this article: what does this primarily consequentialist understanding of accountability imply for those humanitarian actors whose perspectives, including ethical perspectives, on their actions are precisely not consequentialist but, rather, “Dunantist”?
Consequentialism and Dunantism are two key approaches in humanitarian action. While often referred to as two distinct approaches, this precise pair of terms may not necessarily be used.Footnote 3 Consequentialism in normative ethics refers to the approach of justifying and evaluating an action by its consequences or ultimate aim, as opposed to the deontological or Kantian position that what makes an action “right” need not (only) be its consequences.Footnote 4 In humanitarian academic literature, this latter, non-consequentialist position is also termed deontological,Footnote 5 but it is also varied in nuanced ways by referring to value-,Footnote 6 obligation-Footnote 7 or duty-basedFootnote 8 rationality. In these discussions, specific reference to “Dunantist actors”,Footnote 9 who follow the “Dunantist traditions”Footnote 10 and “Dunantist principles”Footnote 11 of neutrality and independence, can often be found. As the use of different terminologies shows, the application of deontological ethics to humanitarian action is not straightforward.Footnote 12 As a determination of the various nuances in the application of deontological ethics in humanitarian action would go beyond the topic of this article, the term “Dunantist” will be used here to describe the non-consequentialist ethical perspectives that these actors embody.
Going back to the application of the concept of accountability, issues of a practical and ethical nature may arise when these Dunantist actors apply the current, primarily consequentialist understanding of accountability. There may be consequences to consider, particularly with regard to the principle of neutrality and the humanitarian access to conflict areas that this principle facilitates. Alternative approaches to the concept of accountability may be preferable for such actors.
To make its argument on the importance of considering the broader implications of how accountability is understood and applied, this article first discusses the concept of accountability and how it has evolved in the humanitarian world over the past two decades. This first section culminates in a reflection on how the concept of accountability, as it is now understood, relates to normative ethics on humanitarian action, specifically regarding Dunantist traditions versus more consequentialist trends. The next section elaborates on issues following the logic of accountability for what, how and to whom, once the selected area of humanitarian protection activities in armed conflict settings is defined. Thereafter, the article expands on the consequences that stem from the application of accountability concepts and argues that accountability can and must still be achieved, only differently.
In conclusion, the paper will suggest that treating accountability from the currently dominant consequentialist, often technical, often developmental, one-size-fits-all perspective is a choice that can be made, but not without practical issues, larger consequences and trade-offs. This should be taken into consideration more seriously, particularly by humanitarian actors who wish to associate themselves with Dunantist ethical perspectives. This discussion can serve as an interesting example of how the current thinking around the humanitarian–development–peace nexus may require deeper reflection by humanitarian policy-makers with regard to the implications and consequences of this development of treating humanitarian and development action as one continuum, as well as potential exceptions and necessary alternative approaches for some actors.Footnote 13
Setting the scene: What is humanitarian accountability?
The concept of accountability
Despite the ubiquity of the term “accountability”, a precise meaning or definition applicable to humanitarian action is evasive. Legalistic understandings, such as whether “parties to a transaction or compact have abided by its terms, performed their respective obligations, or delivered agreed-upon outcomes”,Footnote 14 do not lend themselves well to non-governmental organizations (NGOs) in general, where the relationships between NGOs and donors or affected populations do not typically take the form of binding contracts with clear targets and the possibility of penalties.Footnote 15 Legally speaking, the only accountability that applies may be between employee and employer, or consultants and contracting agencies – for example, for keeping to confidentiality clauses, for agreed-upon deliverables, or for reports and tools, none of which typically include accountability for the humanitarian endeavour as such. The concept of power, rather than legalities, has been used to define accountability, but questions of where power really lies, particularly if one moves beyond contractual relationships to questions of political instrumentalization of humanitarian aid, make this definition largely impractical. Specifically for humanitarian NGOs, Hugo Slim offers a more convoluted definition of accountability in the humanitarian context: “the process by which an NGO holds itself openly responsible for what it believes, what it does, and what it does not do in a way that shows it involving all concerned parties and actively responding to what it learns”.Footnote 16 In this definition, Slim takes accountability a step away from technical aspects and combines it with value-laden concepts of purpose, participation and transparency, and even links it to the concept of a learning organization. Lisa Jordan and Peter van Tuijl conclude that “accountability is a normative and socially constructed concept”, changing over time and open to reinterpretation.Footnote 17
Though a clear definition of accountability is not easily attainable, key tenets of the concept can be identified. First, there is the element of being answerable to someone or some entity, generally for actions and maybe their results and consequences, penalties or sanctions. Given the closeness to the term “responsibility”, there is also an element of accountability to oneself contained in the concept. Second, accountability has an overarching objective in the sense of accountability for something, which is generally performance of a task or aspects of performance such as effectiveness, leading to these terms also sometimes being used interchangeably with accountability.Footnote 18 Third, to show accountability, an element of making accountability apparent or visible must be included.Footnote 19 This is also why accountability is often linked to transparency and reporting.Footnote 20 Finally, there is a question of the overarching purpose of accountability, which is generally the need for a control mechanism, in lieu of legally binding arrangements. These questions of accountability – to whom, for what, why, and how to demonstrate it – will serve as a theoretical framework for further elaboration.
Past influences on accountability in humanitarianism and development
In the 1990s and early 2000s, calls for more accountability and laments about a “lack of accountability” could increasingly be heard, in both development and humanitarianism.Footnote 21 On the humanitarian side, this drive for more accountable and better performance is generally linked to the evaluation of the humanitarian response to the Rwandan crisis in 1994–95.Footnote 22 Both sectors were also influenced by public sector innovations associated with results-oriented management schemes and ideas around “New Public Management”.Footnote 23 Business trends associated with “balanced scorecards” and similar concepts using increasingly available data implied that improved tools for accountability were available.Footnote 24
Between the two sectors of development and humanitarian action, it was probably the latter that was influenced by the former, with development perspectives being increasingly introduced to working in conflict settings.Footnote 25 By 2007, accountability could be labelled a “tenet of humanitarian action”, sitting as equally important alongside principles of neutrality, independence and impartiality, and concepts of dignity, sustainability and participation.Footnote 26 This general acceptance of accountability applies to two distinct meanings of the term: upward accountability to donors, and downward accountability to the affected population. For the humanitarian sector, the Sphere Project,Footnote 27 commencing in 1997, can be seen as an attempt to enable upward accountability through institutionalizing service delivery standards, while the Humanitarian Accountability Project (2002) was developed to enhance downward accountability.Footnote 28
The accountability drive of the past decades links to two more trends, both in humanitarian and development settings. First, there is the increased focus on results and consequences, as in results-based management, and on responsibilities for consequences, or the “broader impact” of humanitarian action.Footnote 29 These concepts link with accountability in the sense of defining what actors should be accountable for: results. The prevailing logic of framing activities in results-based management is linear, with a beginning and an end (inputs–outputs), and then asking about the consequences of activities (outcome–impact). It is now also the prevalent approach in protection activities.Footnote 30 Digitization and “big data” are expected to facilitate the quantitative measurement of such results and impacts.Footnote 31 Second, with regard to questions on how to demonstrate accountability, the use and importance of evaluations has risen. Evaluations allow organizations to offer a critical analysis of achievements to external stakeholders, such as (State) donors or the wider public. For example, the Active Learning Network for Accountability and Performance (ALNAP) has focused on collecting and publicly sharing evaluations.Footnote 32 Again, the ambition with both reporting and evaluations, as with accountability as such, is to be “evidence-based”.Footnote 33
Ethical perspectives on humanitarian accountability
In line with the “rise of the relief-development continuum” that started decades ago,Footnote 34 organizations increasingly interchange between development, relief and humanitarian work, becoming “multi-mandate organizations”, including in armed conflicts.Footnote 35 An influential definition of development is “good change”.Footnote 36 Asking for accountability therefore implies asking for evidence of what “good change” has been achieved. If humanitarian actors consider their activities as developmental, and themselves as “change agents”,Footnote 37 then accounting for change, over a certain period of time, is justified and necessary.
The raison d’être of humanitarian action from this change perspective is seen as the “impact” of an action rather than the “ethical value” of an action.Footnote 38 “Impact” leans towards linear models of thought, including strategic planning, a quest for effectiveness and efficiency, and the comparison between inputs, outputs, outcome and impacts. In contrast, “value” implies that an action is valuable per se, for example as an expression of empathy or sympathy, potentially even regardless of any eventual negative consequences of the action. Eva Wortel describes this differentiation as acting out of a “moral sense of the importance of human life” rather than acting due to a categorical imperative or for utilitarian reasons of achieving a greater good.Footnote 39
A controversial example may be alleviating hunger in situations where violent deaths occur – i.e., the “well-fed dead” that were raised as a key concern in the 1990s.Footnote 40 Simplified, the value-based approach would be to feed hungry people because they are suffering at this point in time, and because the means to do so are available. This does not imply naively disregarding the larger picture of other dangers these hungry people may be facing, but, provided this is feasible given other immediate priorities, alleviating the hunger would be an immediate and unavoidable step to take, largely regardless of its potential effectiveness and efficiency. Impact-based thinking, on the other hand, could be to accept that feeding a population under other deadly threats makes little sense. Instead, prioritizing efforts aimed at the resolution of the larger threats, which may affect more people than those hungry right now, may be the appropriate course of action, even to the extent of potentially ignoring alleviating hunger for this specific group at this same time. This may also include ultimately accepting that – even if the need for food is greatest right here, right now – it is simply a waste to use valuable (food) resources on people who are about to die for other reasons, and the resources have to be better used elsewhere to save the lives of people more likely to survive.
This has parallels to applying normative ethics to humanitarian action, where the terms of consequentialist and non-consequentialist or deontological ethics can be found.Footnote 41 These perspectives have also been called “instrumental rationality” versus “value rationality”Footnote 42 and an “ethic of consequences” versus an “ethic of obligation”.Footnote 43 Some humanitarian actors traditionally align more to one side than the other, resulting in the labels of “Wilsonian” (consequentialist) and “Dunantist” (deontological)Footnote 44 being used, with the International Committee of the Red Cross (ICRC), founded by Henri Dunant, being the key example of the latter.
The current approach to accountability leans in the direction of consequentialist ethics, which are classically more linked to developmental than humanitarian theory but are considered prevalent in today's humanitarian world.Footnote 45 This weakens non-consequentialist concepts such as compassion, which may be “worthwhile” but not ultimately satisfactory, as it likely falls short of changing the situation that causes the need for compassion.Footnote 46 Similarly, non-consequentialist actions tend to have limited ambitions to “only” alleviate rather than eradicate suffering. The overarching reasons for accountability are also given from a consequentialist rather than a Dunantist logic: accountability is ultimately needed to justify the humanitarian action itself, assuming that it can demonstrate its achievement in the sense of results (as in positive change).
Currently, accountability demands on all actors, including Dunantist organizations, are based on the mainstream understanding described above. The International Red Cross and Red Crescent Movement, for example, claims to be committed to the mainstream concept of upward and downward accountability: Article 9 of its Code of Conduct specifies that “[w]e hold ourselves accountable to both those we seek to assist and those from whom we accept resources”.Footnote 47 This commitment to accountability may be limited by a requirement for confidentiality, which is particularly needed for carrying out sensitive protection activities, as will be expanded on in the next sections.Footnote 48 In any case, the ICRC is said to be facing pressure from donors to improve its accountability.Footnote 49
Therefore, the key question should be whether Dunantist actors can, and should want to, fulfil accountability demands that align more to the ethical point of view of consequentialism, and what the consequences of this decision might be with regard to adherence to humanitarian principles. The key principles that cause challenges in this regard are those of neutrality and independence. The objective of this paper is not to engage in a discussion on the necessity of adhering to these principlesFootnote 50 or the related decisions on ethical positions, nor related realities on the ground such as the increasing militarization and politicization of humanitarian action; the objective is, rather, to show that choosing a Dunantist ethical stand, with its stress on neutrality and independence,Footnote 51 entails consequences for the application of the concept of accountability.
Challenges in applying accountability: For what, how, and to whom?
Accounting for humanitarian protection activities in armed conflict settings
It is notoriously challenging to apply the current understanding of accountability to humanitarian protection activities in armed conflict settings, which are generally considered to be “hard to measure”.Footnote 52 Specifically for protection, operationalization of accountability requirements is acknowledged to be “extremely difficult”,Footnote 53 “a long way off”Footnote 54 or even “largely unachievable”.Footnote 55 The focus here is not to conclude that protection actors need to try harder, collect more data, and be more consistent in finding common frameworks.Footnote 56 Rather, this analysis seeks to unpack why it is difficult to account for protection activities, highlights the potential consequences of trying too hard and provides alternative approaches to accountability.
First, it is necessary to take a step back and examine what is meant by “protection activities in conflict settings” for the purposes of this discussion. These types of protection activities are carried out by humanitarian actors not only during armed conflict but also during “unrest, riots, rebellions, uprisings, and other domestic troubles and tensions falling short of war”.Footnote 57 Such situations are fundamentally different from disaster settings, as the violence that is inherent to them is being used by the participating parties. As a consequence, persons affected by the violence might be in need of protection. This kind of protection – from the actions of others – is thus different from protection from natural elements in a disaster setting, even if in practice conflict and disaster may coincide and mix with issues such as State collapse and fragility, resulting in so-called “complex emergencies”.
The second necessary clarification is what is meant by “humanitarian protection activities” themselves. In the ICRC, the term “protection” has traditionally derived from civilians being protected from the use of force in armed conflict. It takes a dignity-based perspective, being “any action, or set of actions, designed to maintain or restore human dignity” in armed conflict, while stressing that related responsibilities lie with the parties to the conflict.Footnote 58 The current mainstream definition of protection is only partly aligned with this, as the following widely used definition, quoted by the ICRC, shows: “all activities aimed at ensuring full respect for the rights of the individual in accordance with the letter and the spirit of the relevant bodies of law, i.e. human rights law, international humanitarian law, and refugee law”.Footnote 59 This definition grounds protection activities not only in international humanitarian law (IHL), and thus in situations of armed conflict, but equally in human rights and refugee law, which also apply outside situations of armed conflict. Referring to these bodies of law broadens the application of protection, now framed in relation to “fundamental rights”. With a rights-based rather than dignity-based definition, protection can include references to social, political, cultural, gender and economic rights, linking it to issues such as democracy, justice, peace and civil society building.Footnote 60 Questions have been asked on whether stretching the concept of protection this far to include issues of social and developmental change has led it to “lose its distinctive meaning”.Footnote 61 Possibly as a result, on a more operational or technical level, a common understanding of what “protective humanitarian action may mean in practice” is seen as lacking.Footnote 62 There may also be a growing disconnect with efforts to counter the concrete threats that hostilities pose to affected populations. Questions around what protection does or does not encompass, starting with the dignity of individuals but now going as far as obtaining social change, will find an echo in the accountability discussion below.
Issues resulting from Dunantist ethical perspectives: Accountable for what?
The currently dominant understanding is that humanitarian actors should be accountable for achieving a desired change or result, taking into account concepts such as effectiveness, efficiency and timeliness. Depending on what protection by humanitarian actors is considered to encompass, achieving change can pose significant issues from a Dunantist ethical perspective.
To understand what change or results might be desired requires going back to the root meaning of the term “protect” itself, and how this relates to the role of humanitarian actors. Protection aims at protecting people from suffering, in the sense of putting an end to violent abuses or preventing them in the first place. This is the raison d’être of IHL, and this duty to protect lies with the parties of the conflict. While the mainstream definition of protection does not provide clarity on whose responsibility it is to offer protection, and insofar as this includes the humanitarian actors themselves, a definition used by the ICRC refers to (State) authorities as those responsibleFootnote 63 and includes other (State) actors intervening with peace-building/keeping mandates.Footnote 64
According to this understanding, protection activities of humanitarian actors are in origin about reminding parties about their duties, rather than aspiring to directly protect persons affected by armed conflicts or other situations of violence. Unfortunately, the more literal meaning that protection actors should protect affected populations by stopping violence has predominated in recent years, and can be found in principal protection handbooks, with an admission that this may indeed be beyond the capacity of most protection actors or activities.Footnote 65 It also appears to be the measuring stick applied by some donors, who ask “what activities can be supported that will effectively afford protection to our affected communities”.Footnote 66 This literal understanding of protection by humanitarian actors is problematic. First, it is problematic simply because it is unrealistic: Marc DuBois goes so far as to call into question the idea that humanitarian actors can actually protect people affected by armed conflict because protection actors and their activities are helpless in the face of purposeful violence and will then be targets themselves.Footnote 67 Actually protecting civilians in an armed conflict is generally only within the power of military forces, implying that accountability of humanitarian actors – provided they are distinct from military actors – cannot be about providing protection in this sense. Following this logic, it has been suggested that military and other political (State) actors should be regarded as part of the overall mechanisms addressing humanitarian needs.Footnote 68 Second, the aforementioned understanding is problematic because it implies a level of involvement in military action that is likely to be in drastic contrast with any ambition to maintaining neutrality.
As explained above, protection activities can include action for social, political, cultural, gender and even economic rights, which can also be problematic from a Dunantist ethical perspective. This understanding implies focusing on the root causes of a humanitarian problem, rather than on addressing symptoms, on making structural changes and being “agents of change” in order to create “just societies”.Footnote 69 If, however, such social change is contested, the pursuit of related goals will clash with the principle of neutrality, in its understanding of explicitly staying away from “controversies of a political, racial, religious or ideological nature”.Footnote 70 This, in turn, raises questions of who decides what the ideal society actually looks like: the donor, the beneficiary, the humanitarian actor, or a different stakeholder such as the government of the country in which protection activities are implemented? It seems fair to say that most humanitarian activities are not only funded by Western State donors and the Western public, but also pursue social values close to the Western model of society, such as the empowerment of women and children, and the importance of a (nuclear) family unit rather than larger kinship concepts. As long as such social values are not objects of dispute, there may be no problems from a Dunantist ethical perspective. Humanitarian actors not wishing to compromise this humanitarian principle – i.e., the Dunantist actors – might need to steer clear of engaging in accountability for achieving social change, despite or because of stakeholders’ expressed agendas. Stakeholders here include not only donors, who come to mind first in accountability discussions, but also parties to the conflict or groups of persons affected and beneficiaries of protection programmes.Footnote 71 In addition, whether humanitarian actors believe in a humanitarian imperative to act versus an understanding of the humanitarian act as voluntary has implications for accountability.Footnote 72 If there is an imperative to act, then the consequences or results of the action are what need to be measured, as the decision whether to act or not is a given. If one regards the humanitarian act as voluntary,Footnote 73 then the actor becomes accountable for the decision to act itself, at each point in time. As Fiona Terry has discussed in detail, this decision to act can have grave implications, and evaluating it is not a technical matter of using management models based on inputs and outputs, but a question of ethical judgements.Footnote 74 Indeed, she argues that the focus on linear accountability tools carries the danger of obscuring the more important discussion around “right” and “wrong” that needs to take place. From a classical Dunantist point of view, the humanitarian act is precisely not imperative, but voluntary.Footnote 75 Going back to the terms that are used to describe a non-consequentialist ethical approach, this shows how terms of duty-, obligation- or value-based rationality have different nuances in their application to humanitarian action.
For humanitarian actors choosing a classic Dunantist ethical approach that is more akin to value-based than obligation-based rationality, being accountable for actually protecting from harm (i.e., through the use of force), for change in the sense of addressing social root causes of suffering and for results following an imperative to act overreaches the purpose of the humanitarian endeavour of alleviating suffering. Indeed, going back to the basics of Dunantist ethics, this overlooks the potential intrinsic value of protection activities independent of any results and consequences, going beyond a need for justification, as a voluntary act of compassion.Footnote 76
Practical issues and their implications: How to be accountable?
The generally accepted means of being accountable, and how to show the worthiness of one's undertakings, is aligned with reporting on results, and this is generally presented in a linear logic of input–output–impact. As a starting point, therefore, two basic key requirements need to be met. First, measurable and quantifiable information is ideally needed to serve as hard data for such evidence. Second, a way to link this quantified impact back to the protection activities undertaken is required. This has already been identified as challenging in many fields in the humanitarian and development sectors,Footnote 77 but there are additional issues specific to protection activities in conflict settings that should be considered.
To begin with the issue of measurability, key problems revolve around protection activities being about social behaviour, and thus being a phenomenon that is complex to measure, due to “perceptual and psychosocial dimensions” of protection.Footnote 78 Even if protection activities are simplified to be understood as being about saving lives, deciding on where and on what kind of project to allocate aid requires a cost-benefit analysis with a monetary value assigned to each life,Footnote 79 which is clearly a challenge. Protection work that is, in principle, not quantifiable includes witnessing, being present, conferring dignity and demonstrating solidarity, or simply being compassionate.Footnote 80 Scientific methods of gathering measurable evidence are a challenge to apply in protection activities: scientific methodologies such as randomized control trials could be performed, for example, as a retrospective case-study-based analysis, but they are yet to be successfully applied in conflict situations, for practical and ethical reasons.Footnote 81
The commonly adopted solution to measurability issues is to refer to indicators. Protection handbooks provide general guidelines on how this can be done, and stress how important a good choice of indicators is.Footnote 82 Realistically, however, finding and practically applying meaningful and comparable sets of standard indicators across contexts and actors is difficult not only due to the diversity of situations, but also because of the frameworks, approaches and activities relevant to protection action. Definitions of protection may be available, but comprehensive methodologies of “common protection problems and related modes of action” that allow for cross-context and cross-time indicators and standards are less so.Footnote 83 Here, the context of protection activities also matters, as activities in armed conflict situations may be subjected to significantly more, and even constant, situational shifts than one-off disasters, making the possibility of unchanging, standardized indicators even within one context unlikely.
Closely linked to this is the widespread lack of available baselines, benchmarks, best practices and data from other actors, or standards of some sort, which are a prerequisite for using indicators for comparative performance measurements that go further than output measurements. Again, this can be particularly challenging in conflict settings due to situational fluidity and the lack of readily available and updated information.
The other standard resource, and approach to accountability, is interviewing or surveying those who should be benefiting from programmes – i.e., what is often called the affected population. This approach also neatly ties into downward accountability towards the affected population. A practical but not insurmountable issue with using such interviews as a source of comparable data is that persons targeted for protection programmes in armed conflict settings may frequently and repeatedly change locations, and indeed situations, in some way or another, over a certain period of time, or access to them may become a challenge. Outside hostilities and armed conflict, one may typically find more settled groups, such as in a rural village or a particular group of persons in a similar situation in an urban setting – for example, pregnant women frequenting a particular clinic. These fluctuations that are more pronounced in armed conflict settings render monitoring a statistically representative group challenging, particularly over a larger time span.
More importantly, however, to undertake interviews, humanitarian actors require not only solid professional skills but also a certain degree of “objectivity” and “distance” to the situation that they themselves, and the persons they are interviewing, are in.Footnote 84 While objectivity may be a challenge in any setting, conducting interviews for the purpose of accountability rather than action can be a particular challenge in the face of the acute suffering that armed conflict can cause. Indeed, the principle of neutrality for humanitarian actors was also developed to counter the natural tendency to take sides and differentiate one's empathy between those considered perpetrators and those considered victims.Footnote 85 Furthermore, listening to beneficiary groups and transmitting their interpretation of events can even endanger the perception of neutrality by stakeholders, as beneficiary groups can hardly be neutral in a conflict that affects them.Footnote 86 Finally, still specifically in armed conflict settings, it is a lot to ask from a victim of violence to step back from his or her experience and to reflect on the performance of agencies – noting, moreover, that these agencies often operate at the margins of victims’ actual problems. To make useful comments on the performance of humanitarian actors, a victim would need to retain a realistic grip on the fact that these agencies are unable to address his/her most pressing needs, such as being free from violence and rectifying past grievances.Footnote 87 To sum up, using data from interviews and surveys of the affected population is thus unlikely to easily serve the purpose of collecting measurable data for accountability purposes, despite the fact that the exercise may have usefulness and be necessary from other perspectives.
A second set of practical challenges for demonstrating accountability, after measurability, are issues associated with attribution and causality. Protection issues are correctly recognized as being “intrinsically linked to external factors” outside the influence of protection actors.Footnote 88 As such, applying a linear model to understanding cause and effect beyond inputs and outputs, already acknowledged to be a challenge in developmental situations,Footnote 89 is even more difficult the more unstable a situation becomes. “Theories of change” have been suggested to address this problem.Footnote 90 This approach may work for those protection actors who are aiming at change and therefore also wish to be accountable for change achieved – i.e., actors following consequentialist ethics. Such tools may also be useful to help Dunantist actors think through the possible implications of an action. However, they are ultimately unsatisfactory as an accountability framework for actions that are in principle not directly aiming at larger (social) change and related root causes, but wish to be accountable for the value of an action at a certain point in time only. Similarly, using theories of change can bring an over-focus on what the humanitarian actors can influence, thus exaggerating attribution and potentially minimizing the ultimate responsibility of the duty-bearer who is using force – i.e., the party to the armed conflict.
To give a practical illustration of dilemmas around measurability and attribution, family reunifications are generally the key objective of protection actors in the context of separated and unaccompanied minors in conflict situations. There is, however, no benchmark to evaluate a process indicator of how many reunifications are facilitated, as this is dependent on the situation, the reasons for separation, and also cultural understandings of family/clan dynamics in the context concerned. Scientific control trials of offering service to one group of children but not the other are ethically unacceptable. Concerning prevention, a fall in new registrations of separated minors is far more likely to be causally related to the evolution of a conflict situation rather than the actions of protection practitioners. It is possible that information on separation, for example if related to child recruitment, will be politically sensitive and manipulated by parties to the conflict to suit their own agendas. Humanitarian actors are therefore unlikely to have full access to assess even such basic information as the actual number of separated children.
Finally, outside the context of child recruitment, proving whether the action of reunification was ultimately best for the child is a value-laden issue in itself: “There might simply be no straightforward answer to the question of whether the intervention [i.e., the reunification] enhanced or decreased the [child]’s well-being.”Footnote 91 This is because the perspective of what “well-being” is may fundamentally differ between children, some of whom may, for example, prefer to remain in foster care, and other stakeholders, who may assume that true well-being can only be offered by biological parents or within the child's own cultural and linguistic context. It may also differ over time, meaning that what caused “well-being” in the short term may turn out to be problematic in the longer term, in a situation where “alternative futures” are impossible to include in the analysis. For example, while a child may be perfectly well in a foster home while growing up, they may start struggling with their unconventional childhood situation as an adult, thus putting into question the original decision.
Evidence-based, consequentialist accountability to donors for this type of intervention is probably impossible, and the often-demanded remedy that humanitarian actors should just “try harder”, as quoted at the start of this section, is not promising. Turning to Dunantist ethics is more satisfying in this situation. From a Dunantist position, there is a need to account for the decision to proceed with the reunification at the specific point in time, in this specific context, for this specific child, based on all information that was available at the time when a decision to engage or not had to be taken. The accountability shifts to the protection actor having taken the “appropriate care and attention” when engaging, in view of the expected outcome, rather than being based on the ultimate outcome itself.Footnote 92
Practical challenges related to the complexities of beneficiary voices and lack of common indicator frameworks are likely here to stay. Hopes that technological advances will provide ultimate solutions to the issues mentioned above are most likely misplaced, and may result in other, new complexities arising.Footnote 93
Accountability to whom? Risks that come with fulfilling accountability externally
Another important, but often overlooked, set of issues associated with demonstrating accountability concerns the inherent requirement of the sharing or publishing of information, making it visible to those who are not the humanitarian actors concerned. The problem, specifically in conflict settings, is that the collection, the possession, the sharing, the publication and the interpretation of information relevant to protection activities likely carries a number of risks, as it will be politically sensitive. As such, there are two risks: first, for the persons who supplied the information, and second, for the humanitarian actor who collects and possesses the information.
In an armed conflict setting, people willing to talk to outsiders are taking certain risks. Risks to the affected population will be amplified in a situation of conflict, simply because “the control of knowledge and self-perception”Footnote 94 matters to parties to the conflict, who therefore might not wish certain information to be shared in any outside sphere. Typical examples are information on people supporting parties to the conflict directly or indirectly, which may be unacceptable for the other party, and mentioning or insinuating practices that are unacceptable or unlawful, such as torture, extrajudiciary detention or killings, or the already mentioned example of child recruitment. Such information may ruin the reputation, and thus the support, for a party to the conflict and may have legal consequences in international criminal prosecution, quite apart from potential military value of the information. Moreover, sharing any information can quickly raise suspicions in certain settings, regardless of what is actually said. These risks may not be visible at the time – they may be hidden from the sight of the practitioners (and the person sharing the information), may be too subtle to be noticed, or may only develop at a later point.Footnote 95 Negative consequences may be more acceptable to the person concerned because assistance or other aid is provided in parallel, but probably less so when the encounter takes place for reasons of collecting evidence for the sake of accountability. Given that the concept of “do no harm” is considered an overarching principle in protection work,Footnote 96 humanitarian actors need to decide how much risk to affected persons is acceptable for the purposes of their accountability.
Next, protection agencies possess information. The treatment and storage of sensitive information needs careful and professional management, primarily in the interests of the affected population.Footnote 97 The protection agency can come under threat for having collected sensitive, protection-related information if a danger of this information being shared is perceived by others. Perception may be as important here as facts.Footnote 98 Retaining a principle of neutrality, if so desired by a protection actor, relies on confidence by all parties that sensitive information will indeed not be shared – including for accountability reasons. “Contextualized information-sharing protocols” – i.e., interagency agreements to share data that oblige, formulize and ensure the sharing between signatoriesFootnote 99 – have been suggested as a response to the scarcity of humanitarian data in conflicts but are unlikely to be the ultimate solution to this issue, as truly ensuring confidentiality across a number of humanitarian actors can be a challenge.
For the sake of accountability, there must typically be public reporting. Balancing this with the need to treat sensitive information with care, maybe even confidentially, can be challenging.Footnote 100 The importance of public reporting to maintain “moral purity”, and hence public support,Footnote 101 may be another reason for a tendency to avoid being overly transparent, coinciding with a reluctance to share potentially deeply sensitive protection information. The simple fact that protection cases may exist can already constitute sensitive data, in the sense that reacting to a protection need proves that there is a protection need in the first place. Protection activities of humanitarian actors related, for example, to child recruitment, violence against civilians, or practices such as arbitrary detention or disappearances are activities directly related to actions prohibited under IHL. In the face of such issues, the parties to the conflict, as well as the international community, will desire to hide or publicize such atrocities. Stakeholders may profit from knowledge about protection issues reaching the public realm, making use of it to influence public opinion, or to justify sanctions, etc. If this happens through advocacy or témoignage, it can be presumed to be intended and serving a purpose, but it can also happen – typically unintentionally – through meeting accountability demands as they are currently understood.
Thus, activity reports of protection actors – often funded by Western States or the general public of these countries – can become politicized. In certain situations, any public reporting, even if only related to activities rather than to observations or impact of activities, can already threaten the factual and perceived neutrality of the actor concerned, from the perspective of the parties to the conflict or other stakeholders. To conclude, while processes of identifying problems, gathering data, drawing conclusions and learning should and must take place within a humanitarian organization, there are reasons why such analysis should be invisible from the “outside”.Footnote 102 Data and analysis of protection problems in armed conflict may indeed need to be kept inside the organization.
In practice, this dilemma between the need for accountability and the need to act responsibly with sensitive information for the sake of affected persons and the humanitarian actors themselves can result in extremely careful choices of words and formulations. To choose an arbitrary but not atypical example, the results can be phrases such as “reports of IHL violations and other abuses were rife”.Footnote 103 This type of phrasing does not actually confirm whether there were violations or not, nor what the humanitarian actor did about it, nor how it did it or with what success, and thus does not answer accountability demands – but it maintains a stance of neutrality.Footnote 104
Given the above-mentioned substantive issues around measuring, attributing, collecting, sharing and reporting on protection activities in conflicts, a final additional danger exists: the more difficult it is to collect and present information in a scientifically robust way, the easier it becomes, through ill intent, negligence or error, to misinterpret or even manipulate information. This applies both to statistical data beyond simple activity reports and to beneficiary voices, particularly when such information is provided to an external audience only selectively in an effort to maintain neutrality and independence. This has already been noted in discussions of a danger of the humanitarian actor himself manipulating or misrepresenting information in an attempt to further the actor's own institutional agenda(s),Footnote 105 but it applies far more generally in conflict settings. where many stakeholders, including the parties to a conflict and the affected population themselves, make use of the voices of humanitarian actors – provided in the context of public accountability – for their own overarching agendas.
Accountability and access: More of one, less of the other?
Given the ethical, practical and risk-related issues that can arise when applying the currently dominant consequentialist understanding of accountability to protection activities in conflict settings, the question is what the consequences of such adherence – despite the above-mentioned issues – may be, specifically for Dunantist actors.
The problem of conforming to the principles of neutrality and independence has already been mentioned. One solution would be to delegate these principles to “second-order status” altogether,Footnote 106 given that they are indeed instrumental principles and have no moral value as such.Footnote 107 However, this should not be done lightly. Their primary value lies in the fact that they aim to enable parties to the conflict to trust that allowing humanitarian actors to operate will not negatively impact their chances of winning the conflict.Footnote 108 This includes practical considerations related to warfare, such as the impact of assistance provided, as well as broader considerations related to social norms present in humanitarian action, to which parties to the conflict may not agree. Accordingly, trust can be lost not only by humanitarian assistance being misused, but also by humanitarian actors – even inadvertently– being seen to be applying, representing or even furthering potentially contested social norms.Footnote 109 Trust can also easily be lost through public “words”, which can include public reporting on accountability: “the activation of ‘voice’ often entails being forced to ‘exit’ the field altogether”.Footnote 110 To stress, perception is as important here as reality.Footnote 111
The consequence of losing the trust and confidence of parties to the conflict is losing what is generally termed “humanitarian access” or “humanitarian space”.Footnote 112 Humanitarian access is seen as necessary to allow the distribution of aid and assistance, and to prevent humanitarian workers becoming targets.Footnote 113 In fact, the need for access goes far beyond this: access is also needed to understand a situation, and to be there and listen to the needs of the affected population.Footnote 114 Access is necessary to ensure, as best as possible, that whatever activities are planned make sense from the perspectives of non-discrimination and impartiality – i.e., to have an approximate overview of needs in order to be able to prioritize the greatest and most urgent of them.Footnote 115 Access may be needed for living and expressing the “humane” and “altruistic impulse” that is part of what defines humanitarian staff.Footnote 116 Access is important to ensure accountability to affected persons, to listen to their voices and to prevent instrumentalization by stakeholders, some of whom will falsely claim needs or atrocities to serve their own – legitimate or illegitimate – agendas.Footnote 117 Without full access, aid and assistance distributions are in danger of not serving humanitarian purposes. If access is sought despite a loss of trust, consequences for the security of humanitarian staff may occur, which is an increasing concern.Footnote 118
Whether or not to engage in acute conflict settings is a choice for many humanitarian actors, particularly the multi-mandate actors.Footnote 119 This choice may be to either engage fully in such conflict settings, with the consequences that this implies for adhering to principles and not fulfilling customary accountability demands, or to focus on working outside these settings with a “wider” (focused on improving people's lives) scope, which does not pose the same accountability-related issues.Footnote 120 In this case, humanitarian principles such as neutrality can, indeed, be delegated to second-order importance, but accountability for how humanitarian actors “will engage the state, the democratic process, local political actors and agendas for transformative social justice” should increase.Footnote 121
It has been argued that any impartial actor has the right to engage in protection activities in crisis and conflict, over and above the ICRC as the classic actor.Footnote 122 While this is correct, the sticking points are precisely the humanitarian principles of neutrality and independence, and how these can be reconciled with reporting demands usually linked to accountability.Footnote 123 Theoretically, humanitarian actors might choose to engage in activities inside and outside conflict settings, as well as some activities being more developmental, thus following a consequentialist ethic, and others more Dunantist, as is the case for many multi-mandate actors.Footnote 124 In practice, however, given that acting in conflict settings requires access granted by parties to the conflict, undertaking humanitarian action in both the wider and the more restrictive sense,Footnote 125 at the same time, could endanger the perception of neutrality that is required to gain and maintain access to all populations affected by conflict.
Alternative ideas: Accountability from a Dunantist ethical position
This cannot sensibly mean that protection activities should escape accountability altogether, even if accountability to donors and the general public, particularly if provided through bilateral or public reporting, imposes specific and fundamental limitations. Accountability can and must be performed, for the sake of basic quality control and learning, as well as from the perspective of responsibility for the funds employed by donors and the access granted by parties to the conflict. Accountability is also necessary in view of the judgement calls that Dunantist humanitarian actors make each time principles such as neutrality are called for: these are difficult judgements, and accountability for them is necessary.Footnote 126
To find alternative ways of thinking about accountability, it may be useful to return to the original framework of accountability: to whom, for what, and how. The key first question is accountability to whom. Here, a few alternative entities come to mind. First, there is the organization itself. To enhance accountability between employee and employer, a strong internal discussion culture, with internal evaluations and audits and related learning loops, will be needed. A potential variation of this could be peer group accountability – i.e., accountability fostered between sub-entities within an organization. This could be – retaining the example of protection activities – between persons responsible for protection activities in different countries, or maybe even within a protection team of a country or context.
Second, there is the protection practitioner himself or herself. The present author undertook a series of interviews with twelve protection practitioners working in conflict settings, mainly for Dunantist humanitarian actors, and these interviews showed that the idea of accountability to one's own conscience ranks high in practitioners’ awareness, often higher in importance than accountability to the organization one is working for and to donors, and on a par with accountability to affected persons. Practitioners stressed their need to be true to their own conscience, precisely because being accountable to outside entities is in practice restricted by issues of risk, associated with the sharing and publication of protection information, and the dangers of manipulation and misrepresentation of information.Footnote 127 Giving conscience a role makes particular sense for a Dunantist approach, as the compassion that such an approach values is, in essence, a personal virtue, while at the same time calling for emotionally restraining it in view of principles such as neutrality.Footnote 128 Here, an interesting parallel is evident with the “personal accountability” of traditional Islamic aid organizations.Footnote 129
Third, even if risks to neutrality may limit reporting and hinder subsequent judgement by outside entities at the time of action, Dunantist actors do not necessarily escape a time-delayed, historical judgement: the historical judgement of the ICRC's decision not to go public with its knowledge of the Holocaust during World War II is a well-known example.Footnote 130 The basis for such retroactive judgement could be formal documentation units and organizational archives, availably internally and open to the public after an appropriate amount of time has elapsed. Unfortunately, not many humanitarian actors maintain such archives, and the ICRC's are probably a grand exception.Footnote 131
As for accountability for what, the aim of protection interventions from a Dunantist tradition has been described as typically being less ambitious: the change of acute symptoms of a protection issue and precisely not its underlying roots in the sense of wider social change, or in other words, “an effort to bring a measure of humanity, always insufficient, into situations that should not exist”.Footnote 132 Taken from this perspective, the change that humanitarian actors should aspire to, and be accountable for, contains reactionary elements: to react to a situation and to do what is possible to address the immediate consequences of the situation, on “presentist” temporal terms.Footnote 133 If the aim is to react to a certain situation, then this point-in-time logic, rather than the linear logic currently prevalent, should be employed for the purposes of accountability: was the right action taken at the right time in a specific situation, and were all necessary components in place at this point in time to enable the action to be taken? This can make particular sense in crisis or conflict settings, with their iterative nature, where performance analysis has been aptly described as resembling a series of “framed photographs” as opposed to the feature-length, continuous film that development can be.Footnote 134 Such an understanding is also a departure from regarding actions as “simply good in and of themselves”,Footnote 135 as is sometimes suggested when talking about Dunantist ethics. The difference is that actions should be good – from the perspective of affected populations, so “beneficial to those who suffer”Footnote 136 – at the point in time when action is taken.
Accountability for a reaction in a point-in-time analysis implies that trying, even if not succeeding, has a certain value.Footnote 137 It means that an actor can only be accountable at an output rather than an impact level, and it avoids the above-mentioned potential clashes with neutrality and independence. It does not mean that the consequences of the action can be disregarded, but the focus shifts toward the point in time of the decision, and what was known at that time, and away from longer-term consequences as they subsequently developed. The duty to know all that could have been known at this time, to prove that one has taken the necessary “care and attention”, becomes of central ethical importance instead.Footnote 138 Despite all care, situations will exist in which humanitarian actions, undertaken with good will, nevertheless increase suffering. The implication that humanitarian actors are not accountable even if their actions increase suffering has been called “painfully paradoxical”.Footnote 139 Indeed, Dunantist practitioners will find that such an understanding of accountability, with adherence to neutrality, is far from a comfortable, easy way out of responsibility, as is sometimes implied,Footnote 140 but actually often leads to intense (internal) discussions, anguish and soul-searching.
Practically, these approaches imply reverting to accountability through less evidence-based means, such as reflection and judgement: more art than science.Footnote 141 To do this, protection actors selecting the path of Dunantist ethics need to invest in strong internal accountability mechanisms to counterbalance the limits of external reporting and evaluations, and in recruiting and training staff to take account of the importance of their maturity, ability for reflective action, and conscientiousness, in order to foster internal accountability. To stress, this is a different facet to the often-demanded, and observed, building of technical expertise and professionalization of recent years,Footnote 142 and is more about an ability for ethical reflection and judgement.Footnote 143 Dunantist actors also need to invest in documenting and specifically archiving their actions and reflections, an aspect that the majority of humanitarian actors – with a few notable exceptions – generally neglect at present. Good documentation of actions, as well as ethical considerations behind an action, will also strengthen academic study and evaluations of humanitarian action, whether from sociological, anthropological or historical perspectives, which will allow more rigorous scientific methods than an organization can currently employ in the context of a necessary monitoring of its actions.Footnote 144 However, accountability in humanitarian action is more than a technical and scientific concept, as this paper has tried to demonstrate in detail.
Conclusion
This paper has elaborated that the current understanding of accountability risks being unsatisfactory, inappropriate and ultimately flawed from a Dunantist ethical perspective. This has been demonstrated specifically with regard to humanitarian protection activities in armed conflict settings. It is unsatisfactory due to many practical issues, such as measurability, attribution and danger of manipulation of data and voices – issues that will be encountered while seeking evidence of results. It can be inappropriate because assessing and accounting for protection activities carries risks for the affected population and for protection actors. Finally, attempts at results-based accountability can ultimately be flawed if protection is considered to be not only about the concrete change that is achieved, but also about the ethical value of the act itself, as a voluntary reaction to a humanitarian situation at a given point in time from the perspective of affected persons.
Fulfilling accountability from a consequentialist perspective means accounting for achieving results, which in turn often implies wider goals than Dunantist actors aim for, such as social change. Such change may be contested, particularly in armed conflict situations, and may therefore put into question the neutrality of an actor, which in turn can limit the access of this actor to the humanitarian space. Therefore, trade-offs need to be discussed within humanitarian organizations, and between them and their donors. For those actors who desire to maintain a Dunantist approach together with adherence to the principles of neutrality and independence, alternative ways of thinking about accountability have been presented. These include accountability for care and attention at the point in time of reaction, evaluated for specific situations through internal audit cultures, through cultivating the conscience and ethical reflections of practitioners, and through committing to historical judgements.
Complexities presented in this paper are specific to humanitarian protection in armed conflict situations. How they apply to other humanitarian activities that are potentially equally difficult to account for, such as prevention activities or advocacy in general, will need further exploration. Some issues, such as attribution and causality, may not be unique to the selected example of protection activities.Footnote 145 Other issues, such as the question of whether to account for the result of an action, are only relevant for acts or actors following Dunantist ethics, since this is not in question for actors applying more consequentialist ethics. The identified risks related to the collection and sharing of protection information are relevant only in conflict situations, and not generally elsewhere.
To quote Jean Pictet, “One must choose.”Footnote 146 Using consequentialist ethical approaches for accountability while maintaining Dunantist ethical perspectives towards one's humanitarian activities is not fully compatible. Doing so may have significant consequences, particularly with regards to trust of others that principles of neutrality are being maintained, and thus access. Accountability is not a technical exercise free of ethical considerations, equally applicable to the full breadth of humanitarian endeavour. The nature and situation of activities matter. Understanding and differentiating each humanitarian actor's or activity's ethical position, and designing accountability accordingly, is a necessary starting point.
