Giving What We Can no longer conducts our own research into charities and cause areas. Instead, we're relying on the work of organisations including J-PAL, GiveWell, and the Open Philanthropy Project, which are in a better position to provide more comprehensive research coverage.
These research reports represent our thinking as of late 2016, and much of the information will be relevant for making decisions about how to donate as effectively as possible. However we are not updating them and the information may therefore be out of date.
Emergency aid refers to help given to populations affected by unpredictable natural disasters or human conflicts. It is often very difficult to calculate the cost-effectiveness of disaster relief because many charities do not provide information about their achievements, it is difficult to do rigorous assessment in a disaster setting, and there is so much variation between each disaster. However, we do not believe that emergency aid charities are typically among the most cost-effective.
On 9 July 2011, after a referendum, South Sudan broke away from Sudan and became an independent nation. In December 2013, fighting erupted as part of a power struggle between President Kiir and members of his cabinet. Around 1.5 million people have been displaced, and many thousands killed since the conflict began.
Typhoon Haiyan was a powerful tropical cyclone that struck southeast Asia in November 2013. It particularly affected the Philippines, and killed at least 6,300 people in that country alone, with 28,689 injured and 1,061 missing as of 17 April 2014. 595,149 houses were damaged and another 489,613 totally destroyed. The humanitarian impact was enormous: huge numbers of people were left homeless, and knock-on effects include the spread of disease, and lack of food, water and medication.
In March 2013, the Seleka rebel coalition launched a coup d’état and overthrew President François Bozizé. The conflict has displaced some 900,000 people, with an estimated 2.5 million people “in need of assistance”. A survey of refugees conducted by Doctors Without Borders found that thousands have died in the conflict, although the exact numbers are unknown as the instability means that both external organisations and government bodies are unable to comprehensively assess the situation.
Since mid-2011 the Horn of Africa has been experiencing a terrible drought following two consecutive failed rainy seasons in 2010 and 2011. Whether through direct impact of the drought or through the influx of refugees, the countries experiencing the worst of the crisis are Djibouti, Ethiopia, Kenya, and Somalia. The impact of the drought and the extraordinarily high levels of malnutrition that have ensued, combined with internal and cross-border population displacement and conflict, has exacerbated physical and food insecurity and resulted in a crisis estimated to involve in excess of 10 million people. The World Health Organization (WHO) reported that Global acute malnutrition (GAM) rates in some areas of Somalia at one point reached nearly 50%, with under-five death rates exceeding 6 per 10,000 per day.
On Friday 11 March 2011, a magnitude 9.0 earthquake occurred 29 km under the ocean near Japan, which caused a devastating tsunami, which in turn led the the meltdown of three nuclear reactors on Japan’s eastern coast. The National Police Agency of Japan estimates that 15,887 people were killed as a result of the earthquake, 6,150 injured and 2,612 missing as of 10 July 2014.
In August 2010, the floods that swept Pakistan reportedly affected 20 million people in 78 districts and killed nearly 2,000 people. As well as this, the floods damaged or destroyed about 2 million homes and 514 health facilities and swamped a land mass the size of England. Before the flood, Pakistan already had about four million internally displaced people and refugees due to the war along the Afghan border and the 2005 earthquake in Khyber-Pakhtunkhwa (formerly North-West Frontier Province). Even before the flood, health indicators in Pakistan were bleak, with maternal mortality at 230 (190–280) per 100,000 live births and under-5 mortality at 89 per 1,000 live births. Most of the populations affected by the floods were comprised of the lowest socioeconomic quintiles and were already facing neglect.
Many will remember the shocking news reports from 12th January 2010, when an earthquake measuring 7.0 on the Richter scale shook Haiti. The epicenter of the earthquake was 17 km from the capital Port-au-Prince, which had a population approximately 2 million. The earthquake ‘inflicted significant damage, particularly to critical infrastructure, including basic utilities, transport, communication, and health’. Many structures collapsed, including hospitals and health centers in the Port-au-Prince area, with heavy loss of staff. One of the most disastrous losses was the destruction of the drinking water and sewer systems. Following the earthquake, immediate health priorities included search and rescue for survivors trapped underneath the rubble, providing surgical/medical services to treat injured survivors, preventing wound infection and providing shelter, food, clean water, and sanitation.
There are a number of reasons to suspect that giving to an emergency appeal may be cost-effective. Health and economic effects of both natural disasters and conflicts disproportionately affect developing countries, and, as a general rule, it is a lot easier to make money go further in these countries. More than 90% of natural disaster-related deaths occur in developing countries, and the losses to natural disasters as a percentage of gross national product (GNP) far exceed those in industrial countries. The reason that the metric of percentage of GNP is important is that an individual’s loss of £5,000 a year from a £20,000 annual income can be worrying, whereas the loss of £50 a month from a £200 monthly income can be catastrophic. Moreover, when there are victims, who, after a disaster, have nothing, it feels like even a small amount will help a lot - a relatively cost-effective intervention.
However, there are also some clear problems with emergency aid, the most important of which is that rapid response times and the need for improvisation means that it is difficult to ensure cost-effectiveness. The Disease Control Priorities Project’s report (DCP2) also points out that “the highly emotional and sensationalized climate of disaster response has long prevented the adoption of a cost-effectiveness approach in decision making.” The international community spends millions on short-term responses to disasters, but there is often not enough funding to ensure even the mid-term survival of victims.
It must be noted that the lack of research in this area means that there is a large degree of uncertainty as to the cost-effectiveness of emergency aid and specific interventions within this area. This may be, at least in part, because of the difficulty of doing research at a disaster scene. Humanitarianism raises ethical questions on the role of observers when action at any cost is expected. Furthermore, there is no time for advance planning and control groups are unavailable, making the research difficult to conduct and lowering the number of academics who will conduct it.
This section is a summary of the first two pages of Improving Effective Surgical Delivery in Humanitarian Disasters: Lessons from Haiti by Kathryn Chu, Christopher Stokes, Miguel Trelles, and Nathan Ford.
After the earthquake, there were an estimated 250,000 injured victims suffering a wide range of injuries. In the first few hours, first-aid and triage stations were vital. Within a few days, operations such as wound debridement and amputations were needed. And in the following weeks, surgical needs actually increased as more patients (often with infected wounds) sought medical care. In the weeks to months after the earthquake, non-trauma surgical needs such as maternity services increasingly needed to be provided by NGOs as the pre-earthquake health system collapsed.
The earthquake in Haiti produced a huge humanitarian response. However, the NGOs involved experienced a huge range of different problems that reduced the effectiveness of their operations. Notably, there were supply delays, a lack of appropriately experienced health professionals, and challenges in coordinating between government, military and non-governmental organisations.
Very few of the over 600 health agencies who responded to the Haiti earthquake had the relevant experience or competence to provide appropriate surgical care, and they often lacked the capacity to build emergency surgical infrastructure. A lack of coordination led to too many agencies trying to provide the same care in the same area, while other sections of Port-au-Prince (Haiti’s capital and largest and worst affected city) were left with no access to emergency care. After a few weeks, many military and humanitarian groups left, leaving behind thousands of post-operative patients. The lack of referral systems or any good communication between NGOs caused problems when patients stopped going to hospitals from groups that had left and moved to other NGOs’ hospitals.
Experienced agencies often have a shortlist of hotspot regions (such as central Africa for conflicts or central America for natural disasters). They then preposition materials at major airfields so that their proximity ensures a rapid response. However, prepositioning requires substantial maintenance costs and a high risk of waste as medical and food supplies expire. What’s more, proximity does not necessarily reduce response times. In the case of Haiti, airport congestion led to large delays getting through the medical equipment that was urgently needed to treat the quarter of a million injured victims of the earthquake. The US military-controlled Port-au-Prince airport gave dignitaries and high-profile media priority, while planes carrying humanitarian supplies were diverted to Santo Domingo (the capital of the neighbouring Dominican Republic), and transported the 200 miles to Port-au-Prince by land.
Finally, specialists (surgeons, anaesthesiologists, traumatologists, emergency medicine doctors) who are experienced in treating war-wounded as well as working in resource-limited settings are very rare. Inexperience in emergency settings can be a hindrance rather than a help. The American College of Surgeons listed hundreds of surgeons to go to Haiti, but agencies did not accept their help due to their inexperience.
These problems are indicative of issues encountered in the majority of emergency aid efforts. For example, poor coordination was criticised in response to the Rwandan genocide and the 2004 Indian Ocean tsunami. In 1991, the UN created the Office for the Coordination of Humanitarian Affairs (OCHA) to improve coordination. However, significant media attention as well as financial support led to an estimated over 2,000 agencies responding to the Haiti earthquake, creating competition and duplication. Smaller NGOs could not function independently and drained UN resources by demanding assistance. In the case of emergency aid it is also quite likely that uncontrollable (from an NGO’s perspective) events will occur that significantly hinder the response - in the case of Haiti it was the airport congestion, but in other cases it could be military blockades or overcrowding. These factors limit the success of aid operations, and cannot be overcome financially, suggesting that in many cases there will be little room for more funding in big disaster campaigns.
In the case of Haiti, almost all major organisations, including the Red Cross, Médecins Sans Frontières, Oxfam, Save the Children and UNICEF raised significantly more than they spent in the first year after the earthquake, once again indicating the lack of room for more funding in this case. However, for a disaster that is less well publicised than most, it is plausible that there will be room for funding in its aid provision organisations, as they will receive fewer donations than the average emergency aid charity. Nevertheless, it is clear that there are some important issues with donating to the largest and probably also smaller emergency aid appeals.
The Disease Control Priorities report assessed the cost-effectiveness certain humanitarian interventions used in emergency aid. We will look at 6 of these.
Search and rescue (SAR) is an important intervention in the first 0-48 hours following a disaster such as an earthquake or tsunami where buildings may have collapsed and people may be trapped. Very few developing countries have the expertise to conduct SAR operations in collapsed multi-storey buildings. As such, SAR teams are often dispatched by industrial nations at high cost. However, SAR has rapidly diminishing returns as people who become trapped often die very quickly without any treatment. SAR teams from other nations take too long to reach the disaster zone, rendering their high cost effectively useless. Nearby SAR teams and local emergency services that have good enough training to conduct SAR operations are the most effective rescue solution in this case.
Field hospitals are one of the most common methods of treating victims of a disaster. However, they often cost too much and arrive too late. In the case of the 2003 Bam earthquake, an estimated $10.5 million was spent on around 10 mobile hospitals, arriving two to five days after the earthquake - by this point almost all casualties had been moved away to other unaffected Iranian provinces. This delay means that field hospitals cannot contribute to immediate trauma care and end up competing for routine care that is already being provided by non-specialist local teams. However, those few mobile hospitals that are equipped to care for non-trauma needs and staying longer than two to three weeks can be particularly effective, although there is no data available on the number of lives saved by these hospitals in any disaster that would not have been saved by local means. The building of simple hospitals to replace older ones is far more cost-effective: in Bam, the cost of rebuilding the entire health infrastructure was $10.75 million, very similar to the amount spent on the mobile hospitals.
One of the worst interventions is that of donating in-kind, for example by sending clothing or tinned food, whether to NGOs or from NGOs to countries. They are of limited use, and, according to the Disease Control Priorities report, “often cause serious logistic, economic, and political problems in the recipient country”. They can divert humanitarian funds from more effective uses to warehousing supplies and building facilities for the safe disposal of pharmaceutical donations.
Disease prevention and control can be relatively effective, if the correct managerial decisions are made. Strengthening existing programs to control the spread of disease can be effective, with the benefits of these programs lasting well beyond the immediate health effects of the crisis. Improving sanitation and public awareness of disease, disposing of waste, and reducing the opportunities of vectors to breed constitute relatively effective methods of disease control (see also WASH, immunisation and malaria). However, other, less cost-effective interventions such as improvised mass immunisations and fogging to remove vectors are carried out all too often.
When huge numbers of homes have been destroyed, for example in a tsunami, or many people have been displaced, for example by war, the creation of shelters is a vital method of ensuring survival. Unfortunately, building enough shelters is commonly achieved through the creation of a tent city. Tent cities are easy to set up, difficult to sustain, and nearly impossible to end. Living conditions are often extremely poor. Sanitation problems cause increased risk of diseases such as cholera. When humanitarian agencies withdraw around a year after a disaster, they leave behind displacement camps in urgent need of funding, reducing their access to water, sanitation and hygiene services. Overcrowding, a lack of policing and a lack of lighting lead to security issues inside tent cities such as increased risk of rape. DCP2 also notes a lack of privacy, loss of family identity and loss of empowerment from those living in tent cities. Distributing construction material or cash subsidies is more cost-effective and more suitable for helping the victims of disasters.
Cash assistance is a plausibly effective method of ensuring a long-term recovery. Although it is difficult for cash assistance to meet immediate life saving needs, it does appear to help with rehabilitation. It is clear that in most disaster-affected countries, those with the largest income will more easily gain access to the services and goods they need, whether that be shelter, medical, food, or anything else. This suggests that income availability is often a limiting factor in long-term recovery, and that therefore cash assistance could speed up this recovery (see also microfinance).
It is mentioned above that, in the case of Haiti, organisations received much more than they spent, at least in the first year. It is often the case that disasters motivate donations above the norm.
The UK Charities Aid Foundation (CAF) regularly surveys the donations of UK adults to spot trends. In 2004, the Indian Ocean earthquake and tsunami sparked a huge number of donations, with 81% of UK adults donating to the campaign, and a total of £392 million being donated. The results of the CAF 2004/2005 survey show that in the time period surveyed there was a mean donation of £170.02 per UK adult per year, totalling £8.2 billion. Of this, 13% was spent on overseas aid - around £1 billion. The CAF calculated that the mean UK adult donation to the tsunami appeal was £5.04, and estimated that £1 of this replaced other giving. This totals around £240 million donated to the tsunami appeal through personal donations, £48 million of which was diverted from other causes. To put this into perspective, 25% of personal overseas charitable donations were to the tsunami appeal, and 20% of those were diverted from other causes. If we assume that these donations would have also been on overseas aid, 5% of the total personal donations budget on overseas aid from the UK was diverted away from other causes to the already over-funded tsunami appeal.
We can make a similar, though less detailed, analysis of the results of the CAF 2009/2010 survey to see the effect of the Haiti earthquake on donations. 48% of UK adults donated to the earthquake appeal, with a total of £107 million donated. The estimated total donations in the 2009/2010 period was £10.6 billion, and the mean donation per donor per month was £31, 24% of which was donated to overseas aid, meaning that roughly 5% of overseas aid donations were to the vastly over-funded Haiti appeal.
With smaller events, the numbers are less staggering, though still surprising. With the 2008 Burma Cyclone, 23% of UK adults donated a total of £19.5m, and in the 2009 Asia-Pacific disasters, 23% UK adults donated a total of £9.3m. In 2011, for the East Africa drought appeal, 18% of UK adults donated a total of £72m. It is clear that donations to crises are disproportionately large, lending credence to the idea that there is little room for more funding in emergency aid appeals.
The cost-effectiveness of emergency aid in general is exceptionally difficult to assess. The wide range of different situations makes the effectiveness of the interventions vary massively depending on the event. However, we will try to make some generalisations about the effectiveness of emergency responses as a whole.
Emergency aid is likely to be rushed and improvised, reducing its effectiveness. The majority of DALYs attributable to disasters occur at the time of the disaster, often before aid organisations arrive. The most cost-effective DALYs to avert are in the mid- to long-term after an event, but aid organisations tend to focus only on the acute consequences of an event. There is also a prevalent lack of coordination across disasters and conflicts, and donating to an organisation that presents a logistical or financial burden, rather than solutions to the immediate problems, will hinder rather than help. The range and likelihood of unpredictable problems such as congestion, overcrowding or blockades that can limit emergency responses and that cannot simply be overcome with extra funding: the biggest challenge in emergency aid is logistics. This, as well as the disproportionate amount of money that disasters receive means that, in a large number of cases, there is little room for more funding. Charities often do not spend as much as they raise, meaning that money donated is likely to be spent on a different intervention. So instead of giving to emergency aid, we recommend giving to a charity where your money is really needed and will make a big difference, such as one of our recommended charities.
Although, in general, we do not recommend giving to any of these charities, we recognise that emergency aid campaigns can be very emotionally compelling, and that the effectiveness of emergency aid varies massively depending on the disaster. In smaller, less well-publicised disasters emergency aid may be relatively cost-effective. We have evaluated 5 different (groups) of charities that have been involved with emergency aid: Médecins Sans Frontières (Doctors Without Borders), Oxfam, the Red Cross and Red Crescent movement, Direct Relief, and Partners In Health. We were looking for cost-effectiveness evaluations of their activities, signs that they follow the suggestions of the Disease Control Priority report, and indications that they do not fall for the most common problems of emergency responses. We recommend giving to Médecins Sans Frontières, if they are active in the area, or, if not, to the local Red Cross or Red Crescent society of the area where the event has taken place or the International Federation of Red Cross and Red Crescent Societies. The charities on which we have focused has been guided in part by GiveWell’s prior work in this area.
MSF were the most impressive charity we evaluated. They are transparent and act sensibly. They recognise that emergency aid operations are often overfunded: for the 2004 Indian Ocean tsunami, they closed their appeal after only a few days, and they did not solicit for funding at all for the Japan earthquake in 2011 despite sending a comprehensive emergency response team. To minimise response times (one of DCP2’s biggest criticisms of emergency aid), MSF has staff in 60 countries at all times.
MSF do not end an operation after only a few weeks, but only when one of four conditions are met: the situation has stabilised (e.g. in a conflict zone), the situation has deteriorated so much that it is now unsafe for its staff, other organisations have the capacity and motivation to restore and develop a medical system that meets the urgent needs of the population, or the most urgent needs of people struck by a disaster have been met and relief operations are no longer needed.
As a case study, MSF has been active in the Central African Republic since 1996, and grows its operation in times of conflict, meaning that it is in a viable position to help before the conflict starts. It also recognises that war wounds are not the most important thing to treat in conflict-based areas, and says that other, not directly conflict related interventions are more important.
Finally, MSF carries out impressive evaluations of its actions and makes all its field research freely available at http://fieldresearch.msf.org. For example, one paper analysed the cost-effectiveness of two MSF surgical trauma centres: Teme Hospital, Nigeria at $172/DALY averted and La Trinité Hospital, Haiti, at $223/DALY averted. They recognised that the most cost-effective hospitals in developing countries can have a cost-effectiveness ratio of around $38/DALY averted, and considered ways to improve their centres. They also wrote the paper on which the above case study on Haiti was based, critiquing their work, and this research corroborated with the DCP2 report in almost all areas.
Oxfam are often considered relatively effective for a large charity. They stop soliciting for funding when they have enough: Oxfam stopped raising money for Haiti when it received approximately $100 million, with $18 million spent in the first six months and a three-year plan to help rebuild. In fact, Oxfam stayed in Haiti for 4 years and launched long-term development programs. Oxfam also stopped raising money for the 2004 Tsunami after raising $225 million.
However, Oxfam’s evaluations were disappointing. Oxfam International produce progress reports, and Oxfam GB produce evaluations of all their operations. Oxfam GB’s “effectiveness reviews” are good at measuring whether or not its individual missions reach certain quality controls, but they generally do not assess cost-effectiveness.[36, 37, 38, 39, 40] The Haiti report for the DEC (Disaster Emergency Council) does assess cost-effectiveness, but only very briefly and not very rigorously, focusing on some cases where money was wasted rather than assessing the actual interventions used. The reports also show routine failure to meet the “Timeliness” quality control, one of MSF’s and DCP’s biggest concerns about the effectiveness of disaster relief. Four of these disasters were less well-publicised, so funding was a limiting factor, meaning that there may be some room for more funding in many cases. However, in the case of the Ethiopia drought response, there were factors of government and NGO inefficiencies in communication, large distances, and dispersed communities which led to its particularly low quality score - communication problems seem significant in many cases and can not easily be overcome by more funding.
The International Red Cross and Red Crescent Movement consists of 189 National Red Cross and Red Crescent Societies run by 13 million volunteers and coordinated by the International Federation of Red Cross and Red Crescent Societies (IFRC). The movement’s system of response and comprehensive evaluations are impressive, as is their leading role in the international humanitarian responses to disasters, although we would recommend MSF over the Red Cross/Crescent movement for larger disasters, and donating to the local National Societies or the IFRC rather than the Red Cross/Crescent Society in your country, as they may take a substantial cut of the donations for themselves.
The IFRC’s Disaster Relief Emergency Fund (DREF) is used to ensure that there is immediate financial support available for National Societies for emergency response to events. The DREF has two possible methods of funding responses: the ‘loan facility’, where start-up finding is given for the IFRC and National Societies to respond to large-scale disasters; and the ‘grant facility’, where funding is given to National Societies for responses to small and medium scale disasters for which no international appeal will be launched or when there is no expected support from elsewhere.
The IFRC evaluations are relatively good. In some cases, such as the 2009 Zambia Floods, the DREF system did not work - a lack of understanding of the system by the National Society led to a delay in the request. However, in most other cases, the DREF system was successful. In the Hurricane Sandy Operation, the Jamaican Red Cross was praised for its excellent coordination with government and other NGOs, as well as its large response, which took some time, but was excellent considering that they were often the only relief organisation available as Hurricane Sandy in Jamaica had been a small disaster in the world media. The response to the Haiti earthquake was disappointing, as there were problems with management and cooperation, as well as insufficient engagement at the highest inter-agency or government levels.
The IFRC reports on the 2013 Typhoon Haiyan and the ongoing Sudan internal conflict show that their responses were exceptional. A rapid IFRC and Philippine Red Cross (PRC) response to Typhoon Haiyan led to dispersed services and support to more than the targeted 100,000 families before the typhoon even struck. There was good cooperation between international and national branches of the movement, as well as a high number of volunteers. The PRC distributed cash support to over 57,000 families - this was aided by the cash programming preparedness which took place prior to the disaster. 50,000 family shelters were constructed, whilst 100,000 families were supported with repair and reconstruction. The report still had criticism: despite the external respect for PRC and IFRC, they were not considered leaders and often acted independently without engaging with the external humanitarian community. The response clearly adheres to some DCP recommendations in low response times, cash distribution, and repairing and reconstructing more permanent shelters.
The Sudanese Red Crescent Society (SRCS) requested an appeal to provide an appropriate and timely response in delivering relief assistance to about 173,000 beneficiaries in the South Kordofan, Blue Nile and Abyei regions after a conflict broke out in 2011 between the Government of Sudan and rebel groups such as SPLM-North. By a government directive, only national NGOs and national staff of international NGOs could be allowed in the operation areas of the Blue Nile and South Kordofan states, which vastly increased the responsibility of SRCS, and therefore increased number of targeted beneficiaries to 300,000 and almost doubled the budget, for which other National Societies helped to raise money. 408,317 people had been reached by the time of the report, and over 100,000 returnees from South Sudan settling in the Blue Nile and South Kordofan states were helped through a different appeal. Plenty has gone wrong and much more needs to be done but SRCS has taken a leading role despite budget gaps and insecurity in the area.
The success of Red Cross/Crescent interventions seems very dependent on the national society: their proximity can mean excellent response times and organised relief, but they vary, and there can also be a lack of coordination or even very slow response times in some cases. The Red Cross system seems likely to be effective at responding to small, less well-publicised disasters due to having a permanent National Society in almost every country. Donations to the better National Societies in less developed countries or to the DREF as ‘pre-emptive’ disaster relief may be fairly effective - this needs further research.
It seems that only one report is available on the website (after a fairly thorough search), although there are broken links to Direct Relief reports on other sites. The one available only lists successes: it does not criticise or make suggestions for improvement. However, their financial transparency is impressive. It does not look as though they pay close attention to DCP or other research recommendations.
Partners In Health only works in certain countries. Its oldest and most comprehensive project is in Haiti, where it is the largest non-governmental health care provider. It has been operating in Haiti since 1998 through its Haitian sister organisation Zanmi Lasante. As such, its disaster response operation in Haiti was good. It mainly focused on long-term support such as infrastructure improvements, rebuilding hospitals, and establishing mid- to long-term support clinics in spontaneous settlements. However, Partners In Health does not seem to take part in short-term emergency aid, and it only responds to disasters in areas where it is active, with its response being part of its long-term missions. For example, with Typhoon Haiyan, Partners In Health did not ask for donations but instead directed people to Médecins Sans Frontières and the Red Cross .
Image credit: Wikipedia
Last updated: July 2014