Dangerous Decisions: On the Ideals and Incentives that Guide Humanitarian Decision-Making


By Johan Eldebo and Leah Zamore

UN Staff Stand Together Marking 2017 World Humanitarian Day. UN Photo/Evan Schneider

Protracted conflicts continue to devastate the lives of many in the Central African Republic and South Sudan. Civilian casualties in Afghanistan are at an all-time high. Humanitarian workers are themselves increasingly the targets of attack. From Syria to the Sahel, UN humanitarian appeals remain substantially underfunded. More than 13 million Congolese require immediate assistance. Globally, more men, women, and children have fled from their homes than at any time since the Second World War, while safer parts of the world increasingly close their borders to them.

These facts — all gleaned from headlines published within the last month — hint at the extent and severity of the challenges facing today’s humanitarian system. How do those on the front lines adapt and respond to such challenges? With humanitarian needs far outstripping supply, determining how and to what ends aid workers prioritize their scarce energies and resources is of the utmost importance. In few sectors is such prioritization more consequential; information more difficult to secure or assess; the external environment more complex and fluid. Yet, while the principles of humanitarian action require that decisions be made impartially and independently, on the basis of needs alone, other factors play an important, and sometimes a decisive, role. At least part of the problem that humanitarian decision-makers face is internal to the system itself.This article explores some of the incentives and challenges facing today’s humanitarian decision-makers by analyzing the tensions between four sets of motivations and goals: competition and compassion; risk and reach of operations; principles and precaution; and significance and success.

The term “decision-making” is used broadly to encompass decisions of significance to an organization, individual, or context. The term “decision-maker” is equally broadly used to include individuals who are in a position with sufficient influence (formal or not) to make such decisions. While not a comprehensive study, and views expressed here are the reflections of the authors, the article is derived from a review of existing literature and 31 interviews conducted in 2017 with staff from humanitarian not-for-profit organizations (NGOs), the United Nations (UN), donors, academics, and independent observers and journalists. These key informants represent a variety of perspectives, including headquarters and field level staff.


There are several narratives about the origins of our current form of humanitarian assistance. A common understanding of humanitarianism is that it was created by, and consists of, people whose goal is to alleviate suffering. As such, the humanitarian system has endeavored to maintain its independence (from governments and political agendas), neutrality (favoring no particular side in a conflict), impartiality (aid is based on need alone), and humanity (aid is motivated by altruism and solidarity rather than profit or self-seeking).

As authors like Michael Barnett, David Rieff, and others have long observed, however, the history of humanitarianism is complex and fraught with tensions and contradictions. For instance, the realization by prosperous states that they could instrumentalize humanitarian action by funding it, and the decision of many aid agencies to agree to that arrangement, has dramatically increased the funding and capacity of the aid sector to hire staff and expand capacity across the globe.

But it has also rendered many humanitarian organizations dependent on a small number of Western donor ministries, often giving the latter outsized influence in the setting of humanitarian agendas and priorities.

Additionally, the increase in resources and coverage has necessitated the creation of large managerial systems and mechanisms. This bureaucratization has made decision-making more structured, systematic, and data-driven, but it has also facilitated extensive reporting to higher-ups regarding how resources are spent and how decisions are made on the ground. The sector is able to act more coherently as a result. But satisfying “the system” has become increasingly important and in some cases an end in itself, while the space for more contextualized, nuanced human decision-making and creativity has shrunk.

These and other developments have helped to produce a culture of decision-making that can often seem like a three-way tug-of-war between the principles of humanitarian action, the rules of the bureaucracy, and the priorities of the donors. The more complex the challenges a humanitarian organization faces, the more urgent, and intractable, the push and pull becomes.

Competition and compassion

While compassion is one of the most important values of humanitarianism, it is balanced by competition for limited funding and positioning. Competition can often be a force for good, as it compels agencies and individuals to be effective in comparison with others. It can encourage innovation and better ways of working. It can attract and motivate high-caliber staff. Yet it also creates a struggle at the heart of humanitarian work.

One consequence of this struggle is that it leads aid organizations to specialize in order to distinguish themselves from others. An organization may choose to pursue a particular type of assistance, such as food aid, or to focus in a particular geographic area, or on a specific population such as women or children.

Here the tug-of-war plays out between the competitive edge for funding and visibility and the compassionate drive to provide whatever is most needed by a given community.

This drive toward specialization has multiple positive elements attached to it. It can foster expertise and allow organizations with different skill sets to cooperate in a complementary manner and together bring their expertise to bear on a specific challenge. It also enables better economies of scale for agencies to respond to and learn from different areas of operation and, where appropriate, to share lessons across geographic areas.

However, it can also result in agencies becoming overly focused on their own “preferred solution.” This drive toward “one-size-fits-all” approaches often aims to maximize income, especially when it is scaled up and replicated. This creates a certain moral hazard, where organizations examine specific local challenges not through the lens of the affected persons’ needs, but instead based on the approaches they are already implementing and with which they have a comparative advantage. Internal strategies often reinforce this tendency by attaching performance indicators to preferred solutions and allocating resources to ensure focus remains on them, creating a powerful incentive for organizations to iterate their preferred solution at the highest possible rate.

The development of a preferred solution, such as a particular methodology to provide healthcare or to empower communities through local advocacy, is often (but not always) driven significantly from headquarters, as many technical experts and decision-makers tend to be based there. A main target audience is often the donor community in capitals, as this relationship is frequently front and center in the competition for funding.

Donor agencies, for their part, are also affected by the tension between compassion and competition. Since most institutional donors are a part of, or are funded by, an elected government, they necessarily exist within the political framework of that government and the historical footprint of the country it represents. Recipient countries tend to receive a significant boost in aid when they become relevant to donor governments’ national security interests and/or are seen as potential countries of origin for massive mixed migration flows. That by extension means that NGOs looking for funding are more likely to pay attention to those same countries and issues.

Donor agencies also face political pressure to ensure that their funds are spent well, on time, and in an uncontroversial way to avoid unfavorable headlines or criticism from their electorates. Such pressure can create bias in donor decision-making that tends to favor NGOs that have a well-established preferred solution, and where scaling up would be comparatively less costly than the development of a new methodology and the creation of the infrastructure to implement it. Innovative solutions are funded, but often not at scale.

Risk and reach

The tension between the desire to minimize risk while maximizing reach creates incentives for decision-makers at many different levels either to remain safely within their systems of compliance, or to take risks to better meet the needs of the most vulnerable. This is especially true for decisions taken in relation to high-risk areas that are difficult to reach and operate in. Both within large agencies and between agencies operating in the same area there is often confusion over which individual or which entity should take what decision, and who is responsible for a decision.

To manage risk and empower decision-makers to support effective aid delivery, the last decades have seen a significant growth in the use of data and evidence to justify, design, and implement effective programs. This has been encouraged by donors who have been calling for humanitarian aid to demonstrate “value for money.” It has also been welcomed by many in implementing agencies who want a better understanding of how exactly their programs impact communities where they work. At its best, this type of data-gathering generates valuable learning about program impact and sustainability. It should be standard practice across the sector.

The flip-side of the data equation is when the value of comprehensive data in decision-making becomes too important, resulting in a dangerously high threshold for action or a high entry cost. It can incentivize decision-makers to not act, or act too slowly, in areas where speed is essential. These areas are also the ones where data-collection is most difficult, and where the contextual complexities inherently challenge the standardized implementation of a preferred solution. Indeed, good data will always lag behind reality, as will humanitarian action that relies too heavily on it.

Building a solid base of evidence in an unsafe country or area, comparable to one you can build in a safe one, may well be impossible or at least significantly more dangerous, expensive, and time-consuming. There is an incentive for data-collectors who want to meet output goals, and decision-makers who want to lower risks, to opt for the safe areas over the unsafe, even though the needs are significantly higher in the latter. Thus we need decision-making approaches that work without complete real-time data, but that build on solid conflict and power analysis and that also combine with solid do-no-harm procedures to continually evaluate the impact of such decisions.

Principle and precaution

Data requirements and compliance mechanisms have significant impact on day-to-day operations. Compliance per se is required and also, rightly, helps make operations accountable. One dynamic that it risks fostering, however, is that implementing staff may feel that reporting requirements are more important than holding themselves accountable to the communities with whom they work. Social accountability initiatives and meaningful community consultations are often in place, but organizations may not always track the power balance at play — with the beneficiary community on one side and the donor agency on the other.

One effect of this imbalance is that it provides humanitarian actors with a disincentive to apply the precautionary principle, which diminishes the sector’s ability to respond to humanitarian needs and live up to humanitarian principles. The precautionary principle may require, for instance, proactively expending resources to prepare for a crisis that may not materialize (e.g., by pre-positioning supplies).

Practically speaking, if the humanitarian community is going to be able to respond effectively to a crisis, anticipated or not, it needs to be prepared to do so. But to be prepared — with plans, good context analysis that include scenario planning, and trained staff — costs money. Stockpiling supplies and pre-positioning staff and resources that risk going unused does not generally constitute value for money by most conventional standards. In other words, preparing to respond more effectively to something that might happen seems to be less valuable per “unit cost” than responding to something that has happened — despite the general agreement that prevention is better than a cure. This systematic miscalculation incentivizes decisions that often come a significant human cost through a culture of structural omission of necessary actions that are often easily foreseeable.

Thus, the behavior incentivized by the value-for-money paradigm can be low-cost to achieve a particular goal in the short-term, but with little attention to its sustainability beyond the agreed time frame, and also with little capacity to address or anticipate any changes in context or unexpected needs. These two shortfalls of evidence-driven interventions, in terms of their long-term sustainability across different contexts, and lack of adaptive space, should be included in donor value for money calculations.

At the moment, there is too little incentive for local managers to prepare for a possible disaster, as they would need to commit resources with no tangible and immediate benefit to their ongoing programs. As a result, the humanitarian community is often faced with incentives to launch large responses only after a famine declaration, or after people are displaced by conflict, with community resilience, preparedness and risk reduction being relegated to secondary priorities. In most cases, these sorts of preventative measures are only possible where donors are certain that the crisis will take place. As a result, humanitarian responses are sometimes late and the preparedness measures insufficient.

That said, there are numerous ongoing initiatives that try to address this pattern, and progress is being made. One of the fundamental challenges is that, while it is generally understood that efforts are needed to build resilience, and that it is necessary to use context analysis to plan for and mitigate the impact of a coming or potential disaster, it is often argued that it is difficult to prove the impact or value for money of this initial investment. An overreliance on, and a preference for, measurable quantitative evidence to prove a negative in this context is a real challenge. While in fact, it is not hard to argue, nor to prove, that having a sold macro-level understanding (which is documented, even if unquantified) of the operating context is not only essential, but also necessary and easily demonstrated as good value for money.

Practically, this can mean that field offices are held back until headquarters or the donor communicates an affirmative decision. Alternately, field offices may choose to implement an unsanctioned decision to, for example, initiate or change an urgent response or relocate staff due to worsening security conditions. Indeed, good humanitarian response sometimes even relies on field managers who actively make unsanctioned decisions to save lives.

The overall outcome is often that decision-makers face strong incentives to avoid risks. This includes a preference to not prepare for shocks, and rather to wait and respond in a reactionary manner to a crisis. Effectively, this element of the humanitarian sector’s culture shifts the risks of inaction to communities in need, and the risk of unsanctioned but necessary action to field level staff. Yet the power to change this culture lies within reach of the humanitarian community.

Success and significance

Much of the above tension is a result of compassionate and hardworking individuals seeking to achieve real significance in a culture that too often rewards success as defined in narrow, donor-driven terms. Indeed, perhaps the primary point of contention among the many motives and incentives facing decision-makers is between short-term measurable success and long-term significance. You can measure how much food you deliver to a village in six months, but you cannot as easily measure in that same time frame efforts that would make that village food secure in the long-term. The same village may well need both.

Individual decision-makers, both those working for donors and implementing-agencies, will generally navigate the system more easily if they opt for tangible and short-term success, rather than promoting innovative or long-term alternatives. Success often has to be achieved, demonstrated, and documented within a given (typically one-year) funding cycle or performance-assessment period. However, significant change takes longer than most funding cycles.

In this light, current efforts toward more comprehensive programming, inter-agency collaboration, and longer-term funding schemes are addressing the right challenges. These efforts are an essential step toward improving the capacity of the humanitarian sector to be more innovative, with the aim of more effectively serving populations affected by disaster and conflict.

Overall, current incentive structures are generally geared towards promoting the status quo of operations, while minimizing the perception that the current model may not be the best. Staff carry out analysis of needs and contexts primarily as they relate to their organization’s capacity or preferred solution (e.g., a needs assessment for food aid). As a result, at both the country and agency level, there is often fragmented analysis characterized by artificially reduced complexity. What is needed is to build better comprehensive national and regional-level analysis, with input from local communities. This helps build humanitarian efforts that lead towards a future where those humanitarian efforts are no longer needed

In the same way, while individual staff and managers may prioritize learning, the humanitarian culture rarely provides direct incentives for staff to learn about the history and broader context of humanitarianism, or new ways of doing their current role that lead to outcomes where humanitarian aid is no longer needed. When individual staff do try to reflect and learn, the system with its often intense and short-term goals needs to become much better at allowing application of such knowledge to encourage improvement and critical thinking. This includes allowing staff to innovate, learn by failure, and try again without fear that a failure to attempt something new is career-limiting.


Decisions for the right reasons
To reach the best decisions, organizations need to look critically at what incentives their decision-makers actually face at the time they make decisions. Once understood, measures can be put in place to ensure that those incentives are appropriate or that incentives not conducive to good decision-making are eliminated or balanced out by others.

Donors also need to understand the connection between how they provide resources and how implementing agencies behave as a result. Some initiatives coming out of the World Humanitarian Summit such as the Grand Bargain are helpful. But more long-term funding that decreases the comparative importance of renewal of funding throughout the implementation period is essential. Overall, donors should consider taking a significant step back from detailed management and influence of programs at the local level and instead ensure agencies are properly equipped to manage programs on their own.

Humanitarian agencies, in turn, need to better understand how funding drives their behavior and find ways to diversify their funding streams or resist undue donor interference. Agencies also need to work together to advocate for more flexible funding models. Agencies here need to ensure their advocacy efforts and their programs come together to form a coherent voice to the donors.

Decisions at all levels can and should be guided by humanitarian principles as well as properly informed by ethics, evidence of a technical nature, and context analysis informed by local communities and staff.

Decisions at the right time
Many bad decisions, and non-decisions, happen because decision-makers are ill-prepared or under-resourced at critical moments. Organizations should ensure that decision-makers can, wherever possible, anticipate upcoming key decisions and ensure that the appropriate preparation to make them well exists. This should include consultation with communities likely to become beneficiaries through timely context analysis, as well as collection of the necessary data and evidence to make a coherent and timely decision on how to act.

Organizations should also establish a practice of identifying when a decision to act is required, and allow for the necessary decision even if the time to acquire the preferred data does not exist, such as in the early stages of a disaster or when one is clearly imminent. The relevant decision-makers should be empowered to act at those points within established and appropriate parameters.

In a culture where sins of commission are punished more often than sins of omission, practice should be established where making a wrong decision for a good reason (in hindsight) is not terminal for a decision-maker’s career, and where accountability for not making a decision is established as well. For example, if a manager acts to prepare for a highly likely disaster that does not happen, he or she should not be punished for having allocated resources towards that end.

Donors, holding significant power, need to enable agencies reliant on their funding to make timely decisions. The self-managed Start Network in the UK currently enables agencies to self-manage resources for both anticipatory and reactive activities — one example of ways to build a good decision-making culture. Patient investment, support, and accountability in such initiatives should be welcomed.

The right decisions
Building on the previous two points, organizations should establish an empowering culture where the right decisions can be made. A key component is to ensure that people entrusted with making decisions have the time, support, and guidance to make them properly. This means establishing a common definition of “decision-making success” that works for (1) decision-makers at the individual level, (2) their organizations, and (3) intended beneficiaries. There should be practical guidance that helps decision-makers weigh and pragmatically document the various incentives at play and how different criteria are judged. This should span the ethical dimension, bureaucratic incentives, input from local communities and staff, security of staff and beneficiaries, and the relative availability of data at the time a decision is required.

Donors should use the power they hold to require would-be recipients to ensure that they have properly consulted the communities where they intend to work and that those consultations and context-analyses ensure their programs will do good in both the short-term and long-term. They should also continuously monitor how their funding mechanisms create incentives and adjust them to make sure they encourage the decisions that are in the best interest of those served by humanitarian aid, namely local communities that face situations of intolerable tragedy.

The authors wish to thank the many people across several agencies and entities consulted for this paper, most of whom did so on the condition of anonymity so none will be named here. The responsibility for any errors and for the content of this paper remains solely with the authors and not with any organization involved in its creation. This article does not claim to reflect nor represent the views of any contributing organization.