In brief 

Humanitarian action aims to do good. But it can also cause harm – a risk that’s poorly understood and poorly measured. Over the study period (2018–2021), the humanitarian system was forced to consider its commitment to ‘do no harm’ more substantively, both in terms of direct risk to individual aid recipients and potential negative impact on conflict and the environment. There was more available evidence about negative impacts than in the last SOHS study period but, overall, it remains poorly scrutinised by the system. 

High-profile scandals gave momentum to the prevention of sexual exploitation, abuse and harassment (PSEAH),490 resulting in new inter-agency mechanisms and a noticeable rise in resourcing. Implementation, however, remained slow and ad hoc. Similarly, new attention to the risks of digital harm was prompted by high-profile cases of data breaches and mishandling, and the system is beginning to learn how to address these. 

Wider attention to humanitarian actors’ long-term engagement in protracted conflict settings – including the addition of peace into the humanitarian–development nexus – increased awareness of how humanitarians engage in conflict sensitivity, influence social cohesion and affect aid dependency. Meanwhile, the rise of climate change on the global policy agenda brought the environmental impacts of humanitarian aid to the fore, with the system expected to make concerted efforts to ‘green’ humanitarian action over the next few years.


Humanitarian aid can deliver both intended and unintended benefits for people in crisis, but it can also have unintended negative impacts on communities and the context in which it operates. The principle of ‘Do No Harm’ refers to a formal set of practices focused on how humanitarian efforts contribute to peace or conflict.491 But ‘do no harm’ (lower case) is also a general ethical principle, enshrined in the Humanitarian Charter, which commits humanitarian agencies to avoid or mitigate negative impacts arising from their work.492  

In the study period for this SOHS, questions around the potential harm caused by humanitarian aid and aid workers were prominent both among aid recipients and in global policy discussions. High-profile scandals around sexual exploitation and abuse by aid workers, which garnered major attention at the end of the last study period, continued to prompt questions about the effectiveness of safeguarding mechanisms. New evidence emerged on the decades-old debate about the potential for humanitarian aid to fuel conflict and create aid dependency in fragile settings; the rise of climate change on the global policy agenda led to greater concerns about humanitarian action’s impact on the environment; and the system engaged in difficult reflections on the harms perpetuated by racial bias. Finally, the COVID-19 pandemic presented new challenges as agencies were forced to weigh the risks of exposing staff and communities to a virus against the potential harms of withdrawal.  

Previous editions of the SOHS provided minimal comment on the longer-term and wider negative impacts of humanitarian aid. This is because there is surprisingly little robust data and evidence on those factors for drawing system-wide conclusions. While more evidence was available on do no harm for this edition of the SOHS, the evidence base remains poor; most humanitarian evaluations fail to engage with these challenges in a substantive way.  

This chapter examines how the humanitarian system addresses direct harm caused to individuals with a focus on PSEAH and digital do no harm, as well as indirect and longer negative impacts on conflict and social cohesion, aid dependency and the environment. The effects of racial inequality within the system are addressed in Chapter 2, while the positive and negative impacts of international actors on local- and community-driven aid are discussed in ‘Focus on: Support beyond the system’ section and Chapter 9.

Are humanitarian actors doing enough to reduce and address sexual exploitation and abuse?

One of the most flagrant harms from humanitarian action is sexual exploitation and abuse committed by aid workers. The 2018 SOHS reported increased attention to preventing sexual exploitation and abuse (PSEA) over 2015–2017, including a new UN strategy on promoting a ‘system-wide approach’ to the issue, but it noted that implementation of PSEA policies and mechanisms remained a significant gap that is rarely addressed in evaluations.493  

Since then, high-profile scandals have revealed the cost of years of inadequate action on sexual exploitation and abuse. The 2018 exposure of sexual abuse in Oxfam’s 2010 Haiti response was followed two years later by reports of widespread sexual abuse by WHO and other agency staff in the 10th DRC Ebola response.494 Simultaneously, the #AidToo movement brought attention to long-running sexual harassment within the humanitarian sector, prompting new high-level awareness of the gaps in PSEA implementation. A wave of resourcing, policy and operational changes followed. 

Attention and resourcing  

Within the UN, steps were taken to implement the 2017 strategy on PSEA.495 This included appointing a Victims’ Rights Advocate and creating a new Office of the Victims’ Rights Advocate (OVRA). On the donor side, DAC established a reference group on PSEA led by Ireland, the UK and Austria, which led to the adoption of a 2019 DAC Recommendation on Ending Sexual Exploitation, Abuse and Harassment.496 Following the Oxfam scandal, the UK government hosted a Safeguarding Summit in 2018, where donors and agencies outlined new commitments and shaped reforms. There were several agency and inter-agency initiatives, notably new IASC implementation plans, a global review of progress on preventing sexual exploitation, abuse and harassment (PSEAH) in the 2018–2021 period, and a number of reviews conducted by Humanitarian Country Teams. 

The study period witnessed a wave of resourcing, policy and operational changes to prevent sexual exploitation and abuse.

Practical initiatives for PSEAH also emerged, such as agency-specific handbooks and tools at country level497 and better inter-agency systems for sharing misconduct information in recruitment processes. The Misconduct Disclosure Scheme, hosted by the Steering Committee for Humanitarian Response (SCHR), was used in over 31,000 recruitments in 2021 and the information it provided prevented 142 potential hires.498 In partnership with UK FCDO, Interpol conceptualised a new system (‘Soteria’) to strengthen collaboration between aid agencies and law enforcement agencies to prevent the hiring of accused or convicted sex offenders.499  

There was also a push towards hiring dedicated PSEAH staff, driven in part by donor compliance requirements. INGOs and UN agencies increased the number of PSEAH focal points and coordinators and shifted resources to improve safeguarding mechanisms – this has been enabled by donors allowing costs for PSEAH implementation to be included in funding agreements. These changes may help address confusion among staff about the responsibilities for investigating SEA complaints; previously designated focal points often acted voluntarily without formal inclusion in job descriptions.500  

A global picture on the extent to which dedicated staffing has improved is difficult to obtain; as noted in the IASC review, it is not possible to confirm how many PSEA coordinators are in post worldwide, due to the ‘fragmented’ way roles are recruited and financially supported.501 Estimates suggest that the number of full-time PSEA inter-agency coordinators in countries with HRPs or refugee response plans (RRPs) more than doubled, from 7 to 19 between 2019 and 2021.502 In DRC, one year on from a technical mission to identify improvements after the Ebola response scandal,503 a regional PSEAH expert noted that six sub-regional coordinator positions for PSEA had been created to support stronger inter-agency coordination and response to PSEAH.504 While many still consider resourcing for PSEAH activities inadequate, long-term PSEAH experts and advocates see recent investments as an important move towards improving practice and culture.505  

Accountability and redress 

It will take time to see the practical effects of improved resources and practices on PSEAH, and recent evidence remains largely negative. This was echoed by the respondents to our aid practitioner survey, the majority of whom rated PSEAH implementation as only ‘Fair’ or ‘Poor’ (60%). Even when PSEAH coordinators are in place, they may lack seniority to drive forward meaningful change or be side-lined when other concerns, such as accessing populations affected by conflict, take priority.506 In some contexts, PSEAH reporting mechanisms are still not designed with an understanding of how communities would prefer to report, leading to low levels of trust and use.507 Hotlines and complaints boxes were still used widely, despite repeated evidence from many contexts that these are not appropriate, given accessibility issues and survivor preferences for face-to-face reporting.508 Survivors perceive reporting mechanisms as creating additional risks; the potential for retribution or social stigma outweighs the likelihood of adequate compensation or seeing their perpetrator face consequences. As an agency-wide evaluation of PSEAH practices noted, ‘assurances to victims/survivors and witnesses regarding their safety and security is often limited, and this is likely to be a significant deterrent to reporting’.509

As an agency-wide evaluation of PSEAH practices noted, ‘assurances to victims/survivors and witnesses regarding their safety and security is often limited, and this is likely to be a significant deterrent to reporting

Follow-up on complaints is a pervasive challenge, both in terms of providing adequate support and compensation to survivors and in holding perpetrators to account through legal processes. IASC agencies are committed to providing ‘survivor/victim-centred’ assistance, but what this amounts to in practice can range widely – from GBV services to livelihoods training and from education to legal support. In many countries, survivor assistance is inadequate and hampered by a lack of dedicated resources or inter-agency mechanisms that can facilitate referrals to PSEAH services.510 An IASC global review found that only a quarter of crisis-affected people would be able to access referrals to sexual and gender-based violence (SGBV) services – and even when they do, often ‘service providers are not able to support all needs of SEA victims, such as extended legal aid, including for paternity cases.’511 Financial compensation for survivors is rare, and there is widespread confusion among humanitarian staff on available forms of support or compensation.512  

Overall, new momentum over the study period shows how change is possible when the system works together at technical and political levels to not only make new commitments, but also design the mechanisms and ways of working needed to deliver on them. However, these efforts have been a long time coming and continue to be ad hoc. Several evaluations and reviews note the lack of progress of pilots and commitments on PSEAH extending back over a decade.513 One UN agency has been working to mainstream PSEAH since 2009, yet still struggled with basic implementation in 2020.514 This slow pace of change raises questions as to why the humanitarian system has taken so long to learn its lessons on PSEAH.515 That it took external media-led investigations to realise long-called-for improvements in how the system deals with sexual assault within its own ranks is an acute reminder of the barriers to genuine accountability in the system and speaks to the power dynamics between crisis-affected people and those attempting to serve them.516

This slow pace of change raises questions as to why the humanitarian system has taken so long to learn its lessons on PSEAH.

Box E: Digital do no harm

Authors: Kristin Bergtora Sandvik, Larissa Fast, Katja Lindskov Jacobsen and Maria Gabrielsen Jumbert517 

Do no digital harm has emerged as an important humanitarian imperative. Crisis-affected populations are often required to give personal data to aid agencies in exchange for assistance and protection, and usually have little say or control over how this is used. Where aid agencies produce large bodies of data on crisis-affected people as part of increasingly data-driven assistance practices, attendant risks and new forms of harm are emerging from how digital data is stored, accessed and shared.  

Several incidents over the reporting period for this SOHS highlighted the multiple dimensions of risk:  

  • Data sharing with authorities. One example of data sharing and digital harm is UNHCR sharing biometric data of Rohingya refugees with authorities in Myanmar, from which these refugees had fled. Such data sharing jeopardised the possibility of a safe return for Rohingya refugees, as the Myanmar regime could re-identify individuals using this biometric data.518  
  • Data accessed by unknown elements. Multiple organisations have experienced data breaches and hacks over the past four years. In 2017, the NGO Red Rose experienced a breach of its digital payment platform, which exposed personal data of those receiving cash transfers. Red Rose later branded the breach an act of ‘industrial espionage’. In 2019, a hack against the UN compromised staff records, health insurance and commercial contract data. In 2022, the International Committee of the Red Cross experienced a cyberattack, compromising its Restoring Family Links services in multiple countries and resulting in the loss of the personal data of more than half a million people.519  
  • Partnerships, uncontrolled data flows and attendant risk. Harm may occur where access to sensitive digital data, collected by humanitarian actors, cannot easily be controlled and where there is a risk that such data may be accessed by non-humanitarian actors. This is particularly an issue where humanitarian actors engage in private sector partnerships, as illustrated by the reaction to WFP’s five-year partnership with Palantir announced in February 2019.520 Palantir is widely known for its role in US counter-terror efforts, including use of its software by the Central Intelligence Agency (CIA), and critics argued that this partnership could potentially entail ‘exploitation of the data in WFP’s “data lake”’, including the biometric data of aid recipients.521 
  • Data, targets and harm. Following the withdrawal of coalition forces from Afghanistan in August 2021, biometric devices – with biometric data collected by coalition forces and other actors – fell into the hands of the Taliban. While not humanitarian data, the case is still indicative of a challenge that could also be relevant in the humanitarian domain: how to prevent abandoned digital data ending up in the wrong hands. Critics have raised concerns about the potential for unintended consequences and harm if biometric databases end up in the hands of actors whose priorities conflict with principles of humanitarian protection. Likewise, the construction of large databases such as PRIME (UNHCR), SCOPE (WFP) and Restoring Family Links (ICRC/Red Cross and Red Crescent Movement) represents an inadvertent but foreseeable creation of targets for harm. Even with data safeguards and diligent management, the amount and type of data available in these databases represents a vulnerability for humanitarian actors.522  

While the past three years have produced examples of different types of digital harm, they have also seen positive developments, including the Signal Code,523 the development of policies on biometrics,524 the development of publicly available guidance on data protection,525 responsible data526 and responsible data sharing between donors and humanitarians.527 Nevertheless, the continued development, uptake and regulatory challenges of new technologies such as artificial intelligence require an emphasis on digital literacy in the sector, as well as ethical reflection about the costs and benefits of digital services in relation to humanitarian principles, and legal and regulatory frameworks.

Does humanitarian aid fuel conflict? 

Social cohesion and localised conflict 

Social tensions can arise if aid is targeted in a way that is perceived to be unfair and untransparent. As seen in Chapter 4, communities and humanitarian agencies often hold different views on who is most ‘in need’ and failure to resolve these disagreements has led to increased resentment and even physical violence across many crisis responses – both among aid recipients and between targeted and non-targeted populations.528 Aid recipients in SOHS focus group discussions said that the lack of transparency on targeting decisions had negative impacts on their community, with some taking action to make assistance more ‘fair’ and reduce social tension by sharing assistance with non-recipients.529  

Aid agencies and host governments increasingly recognise that social tensions around fair targeting are especially sensitive between displaced populations/refugees and host communities. The Rohingya response in Bangladesh offers an example of that tension, including outbreaks of violence, and demonstrates how humanitarian do-no-harm efforts are influenced by the actions of others. Humanitarian agencies attempted to mitigate potential social tensions between the Rohingya and Bangladeshi host community in several ways, such as including host communities in programme planning and targeting discussions. The Bangladeshi government played a strong but not uncontroversial role by requiring aid agencies to offer 25% of their support to host communities in order to retain permission to operate. The government also required programming to be separate for host communities and refugees, which was perceived as exacerbating social tensions.530  

More positively, when social cohesion was included in the objectives of a response, humanitarian activities were able to have a positive impact on inter-group relations. The best examples of this were found in livelihoods programmes: training programmes that included both refugees and host community members led to greater feelings of belonging among refugees and more positive attitudes from host participants towards refugees in Kenya, Lebanon and Turkey.531 The ability of humanitarian actors to positively influence social cohesion, however, faced challenges due to wider social, historical and political dynamics outside their control that proved stronger than direct personal interactions. In two separate livelihoods programmes in Lebanon, Lebanese participants felt more positive towards the Syrians they engaged with through activities, yet their overall views of the Syrian population and competition in the labour market remained unchanged.532  

More positively, when social cohesion was included in the objectives of a response, humanitarian activities were able to have a positive impact on inter-group relations.

In addition to inter-group tension, inappropriately designed humanitarian aid can also increase violence in communities, in particular SGBV. SOHS research in Ethiopia and Bangladesh pointed to increased SGBV in camp settings due to inappropriate shelter and a lack of protection programming; elsewhere, inappropriate latrine design has been linked to increased SGBV risks.533 

Wider conflict dynamics

Humanitarian aid provides an influx of resources to a conflict zone, which can shift incentives in ways that either entrench or reduce conflict. Yet humanitarian agencies and evaluations rarely consider these impacts in detail. Few evaluations in the SOHS synthesis directly assessed the impact of assistance on conflict dynamics, while some flagged the shortfalls in how Do No Harm and conflict sensitivity were being understood and applied by humanitarian agencies.534 The inclusion of peace in the humanitarian–development nexus offers the potential for new thinking and connections on this – though as some commentators note, humanitarian agencies need to also focus on getting the basics right on conflict sensitivity.535  

There are many theories but little hard evidence for the potential relationship between humanitarian aid and conflict. Much of the empirical data comes from academic literature and anecdotal key informant interviews, with mixed findings.536 Previous studies claiming that humanitarian aid had a negative effect on peace were based on data sets dating to 1946 or earlier.537 A systematic review of more recent data, from 2004 onwards, found no evidence that humanitarian aid increased violence.538 Moreover, a 2019 review of humanitarian programmes from 2002–2017 finds that food aid decreased the incidence of civil conflict and the onset and duration of civil conflict, particularly conflict caused by ethnic tensions and weather-driven food insecurity.539 This seems to be supported by the inter-agency evaluation of the 2017–2018 Ethiopia drought response, which found that humanitarian aid reduced pressure on resources and consequently the potential for conflict.540

A 2019 review of humanitarian programmes from 2002–2017 finds that food aid decreased the incidence of civil conflict and the onset and duration of civil conflict, particularly conflict caused by ethnic tensions and weather-driven food insecurity.

The relationship between humanitarian aid and conflict is complex: it is affected by the dynamics play in the conflict, the proximity of aid delivery to zones of active conflict, and the degree to which humanitarian actors have fully considered these in the response strategy.541 

Does humanitarian aid increase dependency and undermine longer-term self-reliance?

In fragile settings where governments are unwilling or unable to meet the needs of their population, humanitarian interventions must strike a balance between supporting people through repeated or prolonged crises and avoiding creating parallel services that undermine local efforts and discourage the development of more sustainable, local services and infrastructure.  

The international aid system can contribute to aid dependency by continuing to directly implement services, while underinvesting in appropriate capacity and administrative support for local actors to lead their own response. Uneven efforts were made by international actors to address this, in connection to localisation commitments (see Chapter 9), as one INGO leader noted: 

We still fail to see that we’re always doing ourselves out of a job, and the more we nationalise [INGOs], the less we’re moving onto a place where you’ve got national civil society addressing its needs and a coherent, local, national government architecture … I think we’ve got some tough questions to answer there.542 

Aid dependency is primarily assessed in humanitarian settings through perception, rather than with clear empirical measures. In several countries where humanitarian agencies have been providing basic services for decades – primarily South Sudan, DRC and CAR – aid dependency is explicitly recognised as a characteristic of the context.543 As discussed in the 'Focus on: Resilience in protracted crisis' section, a majority of aid recipients surveyed by Ground Truth Solutions do not feel that the aid they receive supports them to be self-reliant.544 

At a programmatic level, perceptions of aid dependency vary significantly, even among the same population of aid recipients, and efforts to combat dependency can be against the expressed wishes of aid recipients themselves. A potential example is cash-based assistance: one evaluation documented a decline in the participation of cash payment recipients in livelihoods activities545 while another noted a general concern among humanitarian actors about the potential for long-term cash assistance to create dependency.546 

The humanitarian system also contributes to longer-term aid dependency in other ways, such as by drawing away talented staff from more poorly paid state jobs or by providing higher-quality health services than can be offered by either states or development actors, while failing to build this capacity alongside service delivery.547 Despite widespread recognition of aid dependency in protracted settings, it remains difficult to address, as humanitarians are often constrained by short-term funding cycles or host governments restrictions on the provision of long-term opportunities or support to refugees. 

Despite widespread recognition of aid dependency in protracted settings, it remains difficult to address, as humanitarians are often constrained by short-term funding cycles or host governments restrictions on the provision of long-term opportunities or support to refugees. 

Finally, food aid was found to be a growing concern for aid dependency and as undermining resilience more broadly within the context of climate change.548 In the Sahel and Horn of Africa, climate change is challenging traditional pastoralist practices; here, the influx of food aid may be inhibiting the incentives for and abilities of communities to develop more sustainable and resilient coping strategies.549 Due to challenges in providing timely funding and supporting early action, countries are still primarily dealing with drought crises through short-term emergency responses, despite the existence of alternatives. Despite positive examples – including climate change occupying a central role in how WFP considers and funds resilient approaches food insecurity550 –overall shortfalls in adaptation financing to climate-vulnerable countries551 mean a continued default to humanitarian aid. 

Box F: Doing no harm in Afghanistan

Jennifer Doherty

Humanitarian needs were already rising before the Taliban took control of Kabul in August 2021. The speed of the takeover shocked humanitarians within and outside the country, many of whom anticipated an eventual Taliban takeover months after the removal of US troops and were more immediately concerned with Afghanistan’s looming food security crisis.552 The aftermath of the advance on the capital saw the withdrawal of the majority of international staff and an increase in humanitarian needs across the country as the freeze on international financing and development assistance began to bite. 

Do no harm has been tested in several ways during the initial response to the Afghanistan crisis following the Taliban takeover, demonstrating the trade-offs humanitarians face in attempting to deliver support while mitigating immediate and longer-term harm. The international community – UN agencies, donors and NGOs alike – largely split along two positions on working with the de facto regime, both of which had potential negative consequences for the Afghan population. 

For those that withheld support, cooperating with the de facto regime was viewed as violating do no harm principles due to the Taliban’s record on human rights. The rights of women and girls in particular were a sticking point for many agencies, who wanted assurance that their female staff would be allowed to operate safely before they continued operations in the country.553 On a political level, there were concerns that the stability created by humanitarian aid might legitimise the regime by demonstrating that the country could continue to function under the new de facto government.  

Other agencies put the immediate basic needs of Afghans suffering from hunger above broader human rights concerns that they considered beyond their humanitarian mandate. As one INGO interviewee explained: ‘The way to protect women and children is not to have them dying of hunger.’

The way to protect women and children is not to have them dying of hunger.

Both positions had their costs and required compromises in practice. Concerns over the deterioration of hard-won rights led to delays in aid provision which, coupled with sanctions, left local organisations and staff in a difficult position. They had limited funds coming in to pay their own salaries and they bore the brunt of criticism from communities whose needs were not being met. Local staff expressed concern in interviews that this was eroding trust with communities and adding to already high mental health burdens. 

In the end, and as seen in many difficult conflict responses over the study period, humanitarian organisations opted for access, along the way seeking to secure as much support for humanitarian space and access to services for girls and women as possible. Through coordination meetings at both local and national level, humanitarian actors discussed key principles of humanitarian work with members of the Taliban regime. While working with the de facto government, these actors have sought to maintain space for principles to be upheld and for specific types of programming, such as GBV and peacebuilding activities, to be implemented in the response.554

Is humanitarian action ‘green’?

While attempting to save lives and protect crisis-affected people, the system risks damaging the environment through its presence and wider carbon footprint. Available data on the effects of the system on the environment is limited, with few evaluations considering environmental factors and a lack of consensus among agencies on how they should be measuring their carbon footprint.555 Agencies feel they face trade-offs between addressing people’s immediate physical needs and addressing longer-term environmental concerns.556 However, this period has seen a growing understanding of the system’s impact on the environment and the emergence of stronger initiatives to address negative effects. While practical attention to environmental impacts remained relatively low over the study period, it was growing by late 2021, with several positive examples of shifts in operational policy and more environmentally conscious programming. 

Impacts on the local environment 

Refugee responses with large-scale camps have caused some of the worst documented environmental impacts. One systematic review highlighted the lack of green spaces, limited waste management and low air quality in camp settlements, in addition to illness and deaths caused by polluted water.557 Both in-country interviews and evaluations highlighted the Rohingya refugee response in Bangladesh as a significant environmental threat.558 A key informant in Cox’s Bazar reported that, initially, the response placed pressure on the environment through overtaxing water and sewage systems, waste plastics and traffic, depleted water tables and deforestation caused by refugees burning firewood to cook. Environmental degradation – particularly deforestation – in turn caused tensions with the host population as competition for scarce resources increased. 

At the same time, attention paid by agencies to environmental issues did grow, and some have made positive changes to projects – especially as responses move beyond the initial emergency period. For example, WFP conducted a livelihoods programme focused on reforestation in Lebanon559 and Christian Aid worked to address soil erosion and slope stabilisation in Cox’s Bazar.560 Indeed, anecdotal reports from interviews suggest that Cox’s Bazar looked greener in 2021 than during the environmentally destructive earlier years of the response. In addition to projects with a particular environmental focus, there is emerging evidence that increased use of cash-based programming has positive impacts on the environment by reducing the amount of food flown in from other countries for distribution and the provision of in-kind goods that sometimes go unused.561  

An emerging area of success in the system is increased appropriate and efficient fuel use in programmes, particularly in displacement contexts. For example, in Sudan switching to more fuel-efficient stoves reduced wood consumption in some project locations by up to 50%.562 In Bangladesh, UNHCR and WFP promoted the use of liquified petroleum gas cookstoves to reduce refugees’ reliance on firewood and thereby reduce deforestation.563 

Wider environmental impact 

In contrast to the previous SOHS period four years prior, when climate change and the environment were seen as marginal issues, many agencies have developed policies and hired staff to improve sustainability and reduce environmental harm. In 2020, the UN committed to a new sustainability strategy. Although full implementation will take time, efforts are clearly underway. For example, UNHCR has hired a special advisor on climate change and WFP is developing new policies to ensure environmental sustainability is a consideration for both operations and programmes.564 Other organisations and donors are also moving in the direction of greater sustainability commitments. In advance of the 26th Conference of the Parties to the United Nations Framework Convention on Climate Change (COP26), a Climate and Environmental Charter for Humanitarian Organisations was agreed.565 Initiated by ICRC and IFRC, this commits signatories to prepare for climate change disasters and to reduce their own environmental impacts. By the end of 2021 over 200 organisations had signed. Practical measures are also evident. For example, ICRC, IFRC and the Norwegian Refugee Council are developing greener approaches to logistics, fleets and procurement, while several major donors now require grantees to conform to higher environmental standards.566 A new range of tools and guidance are helping organisations consider and reduce the environmental impacts of their projects. For example, MSF has developed an environmental impact toolkit,567 while the IASC and Sphere have both released environmental guidance.568  

Evidence indicates that these nascent efforts are necessary: evaluations revealed large amounts of carbon emitted by the transport of goods and staff during this period, exacerbated by short-term surge travel.569 A 2019 survey by The New Humanitarian revealed that most surveyed agencies did not count their emissions and among those that did, there was limited comparability in the methods used.570 There has, however, been some progress among agencies in measuring emissions over the longer term. For example, the UN has been tracking emissions across its operations for the past decade, and saw a reduction from per capita emissions of 8.3 tonnes of carbon dioxide (CO2) equivalent in the 2010 edition of the UN’s Greening the Blue report to per capita emissions of 6.5 tonnes CO2 equivalent in 2019.571 They also offset nearly all their carbon emissions, such as through tree planting or solar grids, which is a strategy increasingly used by agencies in the system to compensate for their footprint. The efficacy of offsetting approaches has, however, been questioned including concerns about competition for land use and the lack of incentive they provide for reducing emissions in the first place.572 The full impact of the COVID-19 response on emissions and other environmental effects is yet to emerge. Air travel was reduced by the switch to remote meetings, but at the same time food requirements rose, staff required more vehicles to implement social distancing and large amounts of PPE waste have been generated.573 

While there have been stronger policy and practice commitments made across the humanitarian system – which do at least give concerned individuals a standard against which to hold actors to account – interviewees said that this was not matched with strategic financial commitments. There are some exceptions. For example, WFP has committed funds to better understand the effects of climate change and improve related programming through its Climate and Food Security Analyses and the Climate and Resilience Impact Evaluation window, while the ICRC has announced a Climate and Environment Transition Fund574 to develop its own initiatives in line with the new Climate and Environment Charter for Humanitarian Organisations. It is unclear, however, how many other actors in the system will follow suit – which is problematic given the prediction by aid practitioners and host governments in the 2022 SOHS surveys that climate change is likely to be the biggest external threat facing the humanitarian system in coming years.575


This chapter also refers to PSEAH, which includes not only sexual exploitation and abuse of affected people but also sexual harassment of humanitarian staff. When findings are specific to PSEA (i.e. not sexual harassment of staff) we use this term.


The Do No Harm Project, ‘The “Do No Harm” Framework for Analyzing the Impact of Assistance on Conflict: A Handbook’ (Cambridge: MA: CDA Collaborative Learning Projects, 2004).


Sphere, ‘The Humanitarian Charter’ (Sphere, 2018), sec. 9.; The Do No Harm Project, ‘The “Do No Harm” Framework for Analyzing the Impact of Assistance on Conflict: A Handbook’.; J. Martial Bonis Charancle and E. Lucchi, ‘Incorporating the Principle of “Do No Harm”: How to Take Action without Causing Harm Reflections on a Review of Humanity & Inclusion’s Practices’ (Humanity & Inclusion/F3E, 2018).


ALNAP, ‘The State of the Humanitarian System.’, ALNAP Study (London: ALNAP, 2018), 165.


R. Flummerfelt and Peyton, ‘Power, Poverty, and Aid: The Mix That Fuelled Sex Abuse Claims in Congo’, 2020.


UN ‘Special measures for protection from sexual exploitation and abuse: a new approach Report of the Secretary-General.’ (UN: February 2017. A/17/818).


OECD ‘DAC Recommendation on Ending Sexual Exploitation, Abuse, and Harassment in Development Co-operation and Humanitarian Assistance’, 2019.


Misconduct Disclosure Scheme, ‘Implementation Data. Annual Reporting’, n.d.


Interpol, ‘Soteria. Preventing Individuals from Using Aid Work as Means to Access and Offending against the Vulnerable’ (Interpol, 2021).


T Hanley and et al., ‘Evaluation of WFP’s Capacity to Respond to Emergencies. Evaluation Report Vol I.’ (Rome: WFP, 2019); S. Bond et al., ‘Evaluative Review of UNHCR’s Policies and Procedures on the Prevention of and Response to Sexual Exploitation and Abuse’ (Geneva: UNHCR, 2019); Schenkenberg van Mierop, Wendt and Kabir, Joint Appeal to Rohingya; Broughton and Lee, Protection in Australia’s Disaster Responses.


 IASC, ‘Global Report on Protection From Sexual Exploitation and Abuse And Sexual Harassment’ (Geneva: OCHA, 2021), 51


IASC, ‘IASC PSEA Dashboard’, 2022.


IASC 2020. Senior PSEA Technical Support Mission to the Democratic Congo.


PSEA regional expert.


Global key informant interviews


Bond et al., ‘Evaluative Review of UNHCR’s Policies and Procedures on the Prevention of and Response to Sexual Exploitation and Abuse’, 29. 318;


H. Sultan et al., ‘British Red Cross Bangladesh Population Movement Operation External Evaluation’ (Geneva: ICRC/IFRC, 2019), 48.


This is because complaints/ suggestions boxes are ineffective in populations with low literacy levels, some agency hotlines require mobile phone credit to access, and all hotlines require private access to a mobile phone.


Bond et al., ‘Evaluative Review of UNHCR’s Policies and Procedures on the Prevention of and Response to Sexual Exploitation and Abuse’.


UNOVRA, ‘Mapping of Victims’ Assistance Infographic.’, Infographic (New York: United Nations Victims Rights Advocate, n.d.).


IASC, ‘Global Report on Protection From Sexual Exploitation and Abuse And Sexual Harassment’.


 IASC, ‘Global Report on Protection from Sexual Exploitation and Abuse And Sexual Harassment’. ‘Despite the amount of time the IASC Principals and senior officials have committed to PSEAH, there is little evidence in terms of sustainable structures at field level, interagency investment moving from ad hoc to predictable funding and resourcing or HCs/HCTs/ inter-agency coordinators meaningfully reporting incremental progress against Country Action Plans.’ See also: Steets et al., Drought Response in Ethiopia, 37; J. Baker and I. Elawad, ‘Independent Evaluation of the UNHCR South Sudanese Refugee Response in White Nile State, Sudan (2013 - 2018)’ (Geneva: UNHCR, 2018); Hanley and et al., ‘Evaluation of WFP’s Capacity to Respond to Emergencies. Evaluation Report Vol I.’


Bond et al., ‘Evaluative Review of UNHCR’s Policies and Procedures on the Prevention of and Response to Sexual Exploitation and Abuse’.


IASC, ‘Global Report on Protection From Sexual Exploitation and Abuse And Sexual Harassment’; Baker and Elawad, ‘Independent Evaluation of the UNHCR South Sudanese Refugee Response in White Nile State, Sudan (2013 - 2018)’;WFP, ‘Evaluation of the Joint Programme for Girls Education (JPGE) with Financial Support from the Norwegian Government, July 2014 – October 2017’ (WFP, 2020); Hanley and et al., ‘Evaluation of WFP’s Capacity to Respond to Emergencies. Evaluation Report Vol I.’; Steets et al., Drought Response in Ethiopia; Bond et al., ‘Evaluative Review of UNHCR’s Policies and Procedures on the Prevention of and Response to Sexual Exploitation and Abuse’.


UN, ‘Third Annual Report of the Trust Fund in Support of Victims of Sexual Exploitation and Abuse’ (New York: United Nations, 2020). The report outlines the different activities sponsored by the UN Trust Fund in Support of Victims of Sexual Exploitation and Abuse for people subject to PSEA by UN staff members. The package of different activities included psychosocial support, medical case, income generation (including dressmaking classes) and awareness raising. It did not include reference to justice and accountability of the perpetrator to the survivor within programme activities.


Kristin Bergtora Sandvik (S.J.D Harvard Law School 2008) is a professor of legal sociology at the Faculty of Law, University of Oslo and a Research Professor in Humanitarian Studies at PRIO. She is Principal Investigator on the Do No Harm: Ethical Humanitarian Innovation and Digital Bodies project. 
Larissa Fast is Professor of Humanitarian and Conflict Response and Executive Director of the Humanitarian and Conflict Response Institute at the University of Manchester. She is also a PRIO Global Fellow and a co-Investigator on the Do No Harm: Ethical Humanitarian Innovation and Digital Bodies project.
Katja Lindskov Jacobsen is a senior researcher at the Centre for Military Studies. She is a co-Investigator on the Do No Harm: Ethical Humanitarian Innovation and Digital Bodies project.
Maria Gabrielsen Jumbert is a senior researcher and a research director at PRIO. She is a co-Investigator on the Do No Harm: Ethical Humanitarian Innovation and Digital Bodies project.


On RedRose, see ) L. Cornish, ‘New Security Concerns Raised for RedRose Digital Payment Systems’, Devex, 2018; and B. Parker, ‘Security Lapses at Aid Agency Leave Beneficiary Data at Risk’, The New Humanitarian, 2017,; on the UN hack see. B. Parker, ‘EXCLUSIVE: The Cyber Attack the UN Tried to Keep under Wraps’, 2020; on the ICRC hack, see their press releases. ICRC, ‘Cyber-Attack on ICRC: What We Know’,, 2022.


WFP, ‘Palantir and WFP Partner to Help Transform Global Humanitarian Delivery’, WFP, 2019.


L Raftree, ‘A discussion on WFP-Palantir and the ethics of humanitarian data sharing’, 2019; K. Jacobsen, ‘Biometric data flows and unintended consequences of counter terrorism’, IRRC, 2021, 614.


OCHA, ‘Data Responsibility Guidelines October 2021’ (The Hague: Centre for Humanitarian Data, 2021).


F. Greenwood et al., ‘The Signal Code: A Human Rights Approach to Information During Crisis’ (Harvard Humanitarian Initiative, 2017).


C Nyst, Z. Rahman, and P. Verhaert, ‘Biometrics in the Humanitarian Sector’ (London/Oxford: The Engine Room and Oxfam, 2018).


ICRC, ‘The ICRC Data Protection Framework’ (Geneva: ICRC, 2020).


OCHA, ‘The State of Open Humanitarian Data 2021: Assessing Data Availability Across Humanitarian Crises’ (The Hague: Center for Humanitarian Data, 2021); possibly; IASC, ‘Global Report on Protection From Sexual Exploitation and Abuse And Sexual Harassment’


L. Fast, ‘Data Sharing Between Humanitarian Organizations and Donors: Toward Understanding and Advancing Responsible Practice’, Working Paper (Oslo/Bergen: Norwegian Centre for Humanitarian Studies/NCHS., 2022).


Mansour and Dib Haj, Lebanon Host Communities; H. Audi et al., ‘Lebanon, Livelihoods and Resilience Activities (2016-2019): Evaluation’ (Rome: WFP, 2019).


Daniels, Anderson and Yusuf Ali, 2017 Somalia Response; Solutions Consulting, ‘Final Evaluation Report. Disaster Response in Yemen’ (Dublin/Paris: Concern Worldwide, Agency for Technical Cooperation and Development (ACTED), 2019).


From key informant interviews


Tango International, ‘Evaluation of UNHCR’s Livelihoods Strategies and Approaches’ (Geneva: UNHCR, 2018); Danida and UNHCR, Integrated Solutions Model; H. Roxin et al., ‘Effectiveness of German Development Cooperation in Dealing with Conflict-Induced Forced Migration Crises’ (Bonn: German Institute for Development Evaluation (DEval), 2021).


Audi et al., ‘Lebanon, Livelihoods and Resilience Activities (2016-2019): Evaluation’.


D. Stone and K. Chowdhury, ‘Christian Aid’s Rohingya Crisis Response in Bangladesh’ (Christian Aid, 2019); Baker and Elawad, ‘Independent Evaluation of the UNHCR South Sudanese Refugee Response in White Nile State, Sudan (2013 - 2018)’.


E. Bryld et al., ‘Blind Sides and Soft Spots – An Evaluation of Norway’s Aid Engagement in South Sudan’ (Norad, 2020).


Global key informant interviews


Much of the evidence on whether humanitarian aid fuels conflict comes from country-based research, which remains largely anecdotal, or from the academic literature, which is limited by the fact that much of it looks at the relationship between aid and conflict as a whole, combining humanitarian with development and stabilisation aid.


N. Narang, ‘Humanitarian Assistance and the Duration of Peace after Civil War’, The Journal of Politics 76, no. 2 (2014): 446–60.


S. Mary, ‘A Replication Note on Humanitarian Aid and Violence’, Empirical Economics 62, no. 3 (March 2022): 1465–94.


S. Mary and A.K. Mishra, ‘Humanitarian Food Aid and Civil Conflict’, World Development 126 (February 2020): 104713.


This study also found an overall negative relationship between aid and peace (i.e. aid fuels conflict). However, the research looked at humanitarian aid in combination with stabilisation and other forms of development aid, rather than disaggregated. C. Zurcher, ‘The Folly of “Aid for Stabilisation”’, Third World Quarterly 40, no. 5 (4 May 2019): 839–54.


Global key informant interview.


A Klausen L. et al., ‘Independent Evaluation of the Linkage of Humanitarian Aid and Development Cooperation (Nexus)’ (Bern: Swiss Agency for Development and Cooperation (SDC), 2019).


Tango International, ‘Evaluation of UNHCR’s Livelihoods Strategies and Approaches’.


E.g. N. Crawford et al., ‘The Democratic Republic of Congo’s 10th Ebola Response: Lessons on International Leadership and Coordination’ (London: HPG/ODI, 2021).


WFP, ‘WFP Annual Review 2021’ (Rome: WFP, 2022).


IFRC, ‘World Disasters Report 2020: Come Heat or High Water - Tackling the Humanitarian Impacts of the Climate Crisis Together’ (Geneva: International Federation of Red Cross (IFRC), 2020).


Global key informant interview


Global key informant interview


Global key informant interview


L. Salzenstein and K Pedersen, ‘What’s the Aid Sector’s Carbon Footprint? Everyone Is Measuring Different Things, so We Are Comparing Apples and Pears.’’, The New Humanitarian, 2021.


Stone and Chowdhury, ‘Christian Aid’s Rohingya Crisis Response in Bangladesh’.


 M. Wardeh and R.C. Marques, ‘Sustainability in Refugee Camps: A Systematic Review and Meta-Analysis’, Sustainability 13, no. 14 (9 July 2021): 7686.


Audi et al., ‘Lebanon, Livelihoods and Resilience Activities (2016-2019): Evaluation’.


Stone and Chowdhury, ‘Christian Aid’s Rohingya Crisis Response in Bangladesh’.


S. Brangeon and F. Crowley, ‘Environmental Footprint of Humanitarian Assistance’, Scoping Review (The Inspire Consortium and Groupe URD, 2020).


Baker and Elawad, ‘Independent Evaluation of the UNHCR South Sudanese Refugee Response in White Nile State, Sudan (2013 - 2018)’.


Global key informant interviews.


Global key informant interviews.


Climate Charter, ‘The Climate and Environmental Charter for Humanitarian Organizations’ (Climate Charter, 2022).


These include the UK, Canada, the US, Sweden and Switzerland – see Brangeon and Crowley, Environmental Footprint of Humanitarian Assistance’, Scoping Review


MSF, ‘Environmental Impact Toolkit’ (Geneva: MSF, 2020).


The Environment and Humanitarian Action (EHA), ‘Humanitarian Action and the Environment. Essential Guidance for Humanitarian Actors’ (New York: OCHA and UNEP, n.d.).


N. Mock and H.A Ali, ‘Decentralised Evaluation of UNHCR’s Livelihood Programme in Djibouti (2015-2018)’ (Geneva: UNHCR, 2019); E. Debert, H. Sougato Baroi, and D. Sarkar, ‘External Evaluation. Plan International UK’s DEC Funded Response to the Rohingya Refugee Crisis in Bangladesh’ (Woking/London: Plan International UK, Disasters Emergency Committee (DEC), 2019); Baker et al., Response to Cyclone Idai.


Salzenstein and Pedersen, ‘What’s the Aid Sector’s Carbon Footprint? Everyone Is Measuring Different Things, so We Are Comparing Apples and Pears.’’.


UNEP, ‘Greening the Blue Report. The UN System’s Environmental Footprint and Efforts to Reduce It’ (Geneva: United Nations Environment Programme (UNEP), 2020).


Salzenstein and Pedersen, ‘What’s the Aid Sector’s Carbon Footprint? Everyone Is Measuring Different Things, so We Are Comparing Apples and Pears.’’.


Global key informant interviews.


ICRC, ‘The ICRC Climate and Environment Transition Fund’ (Geneva: ICRC, 2022).


The SOHS aid practitioner and host government surveys asked respondents what they thought would be the biggest threat the humanitarian system would face in coming years. Host government respondents most frequency cited climate change. While locally led action and sufficiency of resources were the top two internal threats to the system for practitioners, the next biggest concern was climate change as an external threat.