A political scientist explains what influences disaster relief aid decisions

germany search & rescue nepal
germany search & rescue nepal

Relief workers from the German NGO International Search and Rescue prepare to head to Nepal with their dog Apache (courtesy of VOA News).

In my last post, I examined a handful of studies that explored the political drivers of disaster relief aid determinations. Two of these studies were from Travis Nelson, a political science professor at the University of Wisconsin Platteville. Unfortunately, Dr. Nelson was too busy with end of the semester tasks to get back to me before I published the original post, but he did send me a couple of thoughts last week. Below are my questions on the research and his responses.

Tim Kovach: In your 2010 article, “Rejecting the gift horse,” you argue that anocratic states are the ones most likely to reject disaster relief aid. Do you think that the Burmese government would have initially rejected international assistance after Cyclone Nargis regardless of the fact that the constitutional referendum was one week later, given its hostile relations with the West? Or was the political transition central to this decision?

Travis Nelson: Having looked at a number of these cases of aid rejection, I think the main common denominator is a fear of perceived weakness. The government of Burma/Myanmar was already weak and facing a popular revolt when Cyclone Nargis struck. Any further perceived weakness could have, from their perspective, further emboldened those efforts. India transitioned from aid recipient to aid donor not necessarily because it no longer needed aid but because it wanted to be perceived as a stronger global power. Even the Bush administration’s refusal of international aid following Hurricane Katrina seems to me to be about perception of strength (to both a domestic and international audience) more than anything else. So my read is that Burma/Myanmar rejection of aid was less about the constitutional referendum and more about these broader concerns. Although the referendum itself was, of course, a factor in these broader concerns.

Tim Kovach: I’m not sure if you have read the 2009 World Bank paper (and subsequent 2011 study) from Guenther Fink & Silvia Raedelli that argues disaster aid is nearly as politicized as official development assistance. If you have, do you have any thoughts on their argument? I know that your study does find some evidence that trade, colonial ties, and military alliances influence humanitarian aid flows, but why do you think that your results differ from some of the other literature on these points?

Travis Nelson: Although I haven’t looked at their study that carefully, I don’t think my findings are that different than those of Fink/Raedelli.  Both of us agree that humanitarian aid is ultimately a mix of political interest and humanitarian need.  Individual donor states differ on which types of “political interest” are a factor in their aid decisions.  But the fact of political interest is consistent.  The same is true of humanitarian need.  When I give presentations about humanitarian aid, I often find that audiences are deeply skeptical that humanitarian aid is about anything other than the political interest of the donor.  So I think it is notable that the data does not support this conclusion.  There is still at least some humanitarianism in humanitarian aid.


Here’s how oil, population, and trade affect disaster aid flows

Damage in the Irrawaddy Delta after Cyclone Nargis hit Burma on May 2, 2008 (courtesy of OCHA).

Damage in the Irrawaddy Delta after Cyclone Nargis hit Burma on May 2, 2008 (courtesy of OCHA).

Earlier this month, I wrote a piece for Vox that examined how media coverage of certain natural disasters – or the lack thereof – can significantly affect both the likelihood of a country getting relief assistance and, if it does, the amount it receives.

I don’t want to leave readers with the impression that media coverage is the only, or even the primary factor driving disaster aid decisions; far from it. In fact, there is a fair amount of research that shows how political considerations may be the key issue dictating aid considerations.

Disaster relief is a two-way street

One important factor to consider, as I alluded to in my Vox piece, is that relief aid decisions are a two-way street. Just as donor countries determine whether or not they want to provide support, affected countries can also control whether or not they request it. Without this formal request, the United Nations cannot issue a Flash Appeal, and donor countries will have no way of getting their financial, logistical, and human resources on the ground.

So why would a country hit hard by a disaster actually refuse to accept help? In a 2010 study (gated), Travis Nelson, a political science professor at the University of Wisconsin-Platteville, argued that, for countries transitioning from autocratic systems to more democratic ones – known as anocracies in political science parlance – “aid refusal…is at its core a political act” meant to show that the country’s leadership is able to handle the disaster recovery process on its own.

A recent example of this phenomenon happened in Burma, which I detailed in a 2013 paper (PDF). Cyclone Nargis struck the southern portion of the country in May 2008, killing more than 138,000 people. But rather than accepting offers of support from the international community, the country’s ruling military junta went to great lengths to keep the humanitarian community out. It waited several days to accept international assistance, and even then it only did so on the condition that it strictly control every aspect of the process. It took a full 3 weeks for then-leader Than Shwe to allow the international community to conduct a relief and recovery effort.

Why Burma’s leaders rejected humanitarian assistance

If you view this episode in a vacuum, the junta’s actions make no sense. Why on earth would they actively refuse support, given their clear inability to help their own people?

But, when placed in context, as Nelson argues in his study, things become clearer. Burma, which had existed on the fringes of the international community for nearly two decades at that point, had valid reasons to be wary of offers of support from Western states. These same states had spent years actively undermining and isolating the junta. Moreover, Western leaders took a number of provocative steps in the days after the cyclone that likely delayed the response effort further. The United States anchored a Navy battleship just miles off the Burmese coast, and French Foreign Minister Bernard Kouchner threatened to invoke the principle of the Responsibility to Protect as a means to forcibly initiate a response effort.

One can almost – almost – forgive the junta for fearing that Western governments might use humanitarian actions as a cover for furthering regime change. As Burma expert Andrew Selth has written, “Even paranoids have enemies.”

It’s important to point out that, in accordance with Nelson’s study, the Burmese government had recently initiated a political reform process. Nargis hit just one week before a long-anticipated referendum on the country’s new constitution, a critical part in the junta’s years-long process to slowly and partially democratize the country.

I should also note that, regardless of how irrational the junta’s fears of foreign influence may have been, they actual proved to be somewhat prophetic. Thein Sein, the country’s current president, headed up the regime’s response effort and worked closely with the United Nations and Association of Southeast Asian Nations (ASEAN). This positive interaction between Sein and his international counterparts likely hastened the country’s political transition.

The politics of disaster aid

But now we need to consider the political concerns behind the other side of the equation – donor decisions. What accounts for the fact that the US provides relief after less than 1 out of every 5 disasters?

In a 2009 working paper for the World Bank1, Guenther Fink, an international health economics professor at the Harvard School of Public Health, and Silvia Redaelli, a senior economist at the Bank, explored these factors.

As the authors note, humanitarian aid is often seen as a different animal than official development assistance (ODA). Whereas donor countries have always attached political conditions to ODA, that rule has not applied for emergency aid. Instead, “donor governments perceive emergency aid as political unconditional.” In principle, it should be apportioned according to the greatest need.

But, as Fink and Redaelli demonstrate, donor countries have consistently failed to live up to these principles. In the study, they explore the factors driving the delivery and amount of aid after 449 rapid-onset disasters. They explored the amount of aid provided in the aftermath of these events based on information from the UN Office for the Coordination of Humanitarian Affair’s (OCHA) Financial Tracking System. To control for political considerations, they add a number of independent variables, including population size, GDP per capita, Freedom House rankings, trade openness, the distance between donor and recipient state capitals, whether or not the recipient country exports oil, and how closely the donor and recipient country’s UN voting patterns line up.

Population density, oil exports, and trade openness all affect aid

They find that both the number of people affected and killed increases the likelihood of aid delivery, by 10-13% and 20%, respectively. But, more than that, political factors seem to drive the decision making process. Donors are 25-30% more likely to support their former colonies, and Western countries are far more eager to support oil producers.

Interestingly, donor countries appear more apt to contribute to disaster-affected countries that are not already aligned with their political interests. Unlike with ODA, which donors seem to use to reward their allies, humanitarian aid seems to be a tool to build better political relations. As the authors note, “If the acquisition of international consensus is on donors’ political agenda, emergency aid may well be a more visible, cheaper and more flexible tool to reach such a consensus than traditional development assistance.”

These political issues also mediate the amount of aid provided. When disasters occur at least 1,000 kilometers away, aid values are halved. In contrast, politically non-aligned states see 200% more aid, while former colonies can expect to receive 5 times more.

The correlation between the current allocation of aid and the actual humanitarian losses associated with natural disasters is surprisingly low.

But Fink and Redaelli dig further into the data and find that the five major donor countries – the US, Germany, Japan, Norway, and the United Kingdom – are all motivated by different interests. While the US tends to support democratic regimes, Norway is actually 14% more likely to support more autocratic countries. Germany, for its part, is 66% more inclined to provide aid to neighboring states, while Japan shows an interesting affinity for densely populated countries. And, reminding us all that the sun never set on the British Empire once, former colonial bonds increase the UK’s aid probability by nearly 30%.

The study includes two additional findings. First, the US, UK, and Norway are more likely to support oil exporters by 24%, 35%, and 39%, respectively. Secondly, the authors note evidence for a “bandwagon” effect – when more of these big 5 donors contribute to the relief effort, the total number of donor countries increases significantly.

Unlike other research, the paper also explores what factors drive private disaster aid flows. Nongovernmental donors appear more inclined to support poorer countries and are 16% less likely to provide aid to countries that get high marks for trade openness. Interestingly, private donors do not seem responsive to either the number of fatalities or whether the disaster-affected country exports oil.

Ultimately, as Fink and Redaelli conclude, “While the evidence of the various biases varies significantly across countries, the correlation between the current allocation of aid and the actual humanitarian losses associated with natural disasters is surprisingly low.”

I should note that note everyone agrees with this thesis. Nelson, for his part, authored a 2012 study (gated) that found while these types of political variables do play a role in aid discussion, “humanitarian variables are consistently significant predictors of disaster aid provision.”

So while this debate may not be entirely settled, the fact remains that the international community has a long way to go before it meets the principles laid out in the 2005 Paris Declaration on Aid Effectiveness and reiterated the 2011 Busan Partnership. If we hope to shift from a donor-recipient relationship to a true partnership, developed countries will need to more closely balance their underlying political considerations when apportioning disaster aid.

* Dr. Fink published a version of this paper (gated) in the May 2011 edition of the journal World Development.

The political drivers of disaster in Burma

Recently, Dr. Ilan Kelman of the Center for International Climate and Environmental Research in Oslo posted my paper on the political dynamics of Cyclone Nargis on his website about disaster diplomacy. I appreciate his willingness to publish my work, and I hope it contributes to the development of disaster diplomacy as a field.

Cyclone Mahasen as it looked, forming in the Bay of Bengal on May 12.

Cyclone Mahasen as it looked, forming in the Bay of Bengal on May 12 (courtesy of CNN iReport).

Coincidentally, the publication of the paper happened just as Cyclone Mahasen was coming ashore from the Bay of Bengal into eastern Bangladesh and western Burma. The center of the storm was located near Chittagong, Bangladesh’s second largest city. Fortunately, the storm weakened from a category 1 cyclone to a tropical storm before it made landfall, reducing its potential impact. That said, even the weakened storm killed at least 75 people in the affected areas, rendered thousands of people homeless, and damaged crops. Its impact was clearly less devastating than other recent cyclones, including Nargis (which killed more than 130,000 Burmese in May 2008) and Sidr (which killed 3,500 people in Bangladesh during 2007, after passing through the Sundarbans Mangrove Forest).

Similar to Nargis, however, Cyclone Mahasen highlighted the ongoing political rifts within Burma. In this patchwork state, composed of more than 100 ethnic groups lumped together by the British, ethnic hostilities and rebellions have been endemic. The Uppsala Conflict Encyclopedia lists at least 10 different conflicts between separatist ethnic groups and the Burmese government.

Fortunately, the new government in Burma, which replaced the ruling military junta in March 2011 has worked to settle many of these conflicts, signing a rash of ceasefires and peace agreements. One lingering ethnic challenge which has actually gotten worse since this point, however, involves the minority Rohingya people. The Rohingya, a group comprised of at least one million ethnic Muslims, have experienced severe and ongoing repression within Burma.

The Burmese constitution, which supposedly celebrates the country’s ethnic diversity, does not acknowledge the Rohingya. Despite the fact that at least 800,000 Rohingya live in Rakhine state, they are officially stateless persons, according to the Burmese government. Around 140,000-200,000 Rohingya live in displacement camps in Rakhine, where they struggle to provide for their basic needs. Additionally, they live under constant fear of violent repression by radical Buddhists in Burma, who view them as invaders and threats to Burmese identity. Many of the same monks who were trumpeted by the West for their role standing up to the junta in the 2007 Saffron Revolution have denounced Rohingya and/or led attacks against them.

A young Rohingya girl and her sibling brace for Cyclone Mahasen in an IDP camp in Rakhine state (courtesy of The Los Angeles Times).

A young Rohingya girl and her sibling brace for Cyclone Mahasen in an IDP camp in Rakhine state (courtesy of the Los Angeles Times).

Given this recent history of ethnic violence, it is unsurprising that many Rohingya resisted efforts to relocate them in the wake of Cyclone Mahasen. Fortunately, the storm appears to have largely spared Rakhine and will leave nothing like the trail of devastation from Nargis. However, the disaster and the political issues surrounding it did generate casualties. At least 100 Rohingya crammed onto small boats in the Bay of Bengal as the storm approached, attempting to find refuge in Bangladesh (which refuses to provide asylum to most Rohingya). Tragically, several of these boats sank off the Rakhine coast in the days before the storm, killing an unknown number of people.

This story demonstrates, once again, the fact that disasters are far from “natural” events. As Brad Adams, the Asia director at Human Rights Watch, told the BBC, “any disaster that results [from Mahasen] will not be natural but man-made.” The vulnerability and physical exposure of Rohingya people to Cyclone Mahasen were the products of the social and political environments. They were intentionally constructed by people in power.

This “myth of naturalness” creates the impression that disasters are all acts of God that are beyond human capacity to manage. This perception is largely false and provides cover for malevolent actors, like the military junta in Burma, to deflect from their own actions. Disaster are political events, and they demand political responses. The Burmese people cannot hope to enjoy peace, stability, and development if its state continues to formally repress and marginalize the Rohingya.