Humanitarian crises: testing the CNN-effect

"Lucky are the people of Yugoslavia and Somalia as the world's eyes rest on them. Condemned are the people of Juba … It may be a blessing to die in front of a camera - then at least the world will get to know about it. But it is painful to die or be killed, without anybody knowing it."
Hand-written letter smuggled out from the besieged Southern Sudanese town of Juba, August 1992.

What determines the level of emergency assistance? This paper examines the hypothesis that the volume of emergency assistance generated by any humanitarian crisis is determined by three main factors working either in conjunction or individually. Quantitative and qualitative indicators from recent and ongoing humanitarian crises - both natural and conflict-induced - are used to ask whether the amount of emergency assistance depends on the intensity of media coverage, the degree of political/strategic interest of donor states and/or the level of stakeholder commitment (the strength of the network of humanitarian organisations operating on the ground).

The crises selected for analysis and comparison were selected for their diversity and their individual ability to substantiate each of the three competing explanations. Data has been gathered from the two major TV channels in Denmark and 23 leading newspapers in six states of the European Union and the US. OCHA and ECHO databases were used to approximately identify aid allocations.

Three determing factors

The link between media attention and political action - the 'CNN-effect'- suggests that the media, particularly television, influence the decisions and foreign policy agendas of Western governments. It is commonly assumed that massive media coverage of a humanitarian crisis will lead to increased allocations of emergency funds and improve the prospects of meeting humanitarian needs.

Academic literature, however, provides no substantial confirmation of the existence of a CNN-effect. Studies indicate that the media have had an effect only in situations where the governments involved were lacking a clear policy. Cited 'proofs' of the massive influence of the media are the interventions in Iraqi Kurdistan in early 1991 and Somalia in December 1992. These seem to be exceptions to the rule since media coverage of human suffering only rarely leads to Western policy initiatives. Rather, there is a general tendency for politicians and governments to turn the media into their 'servants' by communicating the message of the government to the public.

If the mass media choose to focus massively on a crisis, a number of preconditions have to be fulfilled. The crisis has to be newsworthy and generate dramatic and emotive footage and stills. Moreover, a humanitarian crisis - particularly an African one - has to compete with emergencies in other parts of the world. Editors do not perceive it as 'news' when Africans kill each other. Neither is it 'news' for most mainstream media if Africa experiences yet another humanitarian disaster. The 'news attention cycle' is another factor; people in the North only pay attention to crises affecting distant people on a cyclical basis and rarely have an appetite for more than one crisis at a time.

A second possible explanation for the level of emergency assistance relates to the interests, especially security concerns, of donor governments. We may assume that emergency donors are basically motivated by the same kind of interests as when they grant long-term development assistance. These include security, trade, investment and wider political interests. According to the 'aid motivation literature', the allocation of development aid from major donor states tends to be motivated by their own national interests whereas donors such as the Scandinavians are mainly motivated by the needs of the recipients. A basic assumption of the donor interest explanation is that the amount of aid received by any low-income country is proportional to the level of interest of the donor. This paper assumes that the same relationship can be found in relation to emergency assistance: that donor interests play an important role in motivating decisions on granting aid to specific humanitarian crises.

As for the significance of stakeholder commitment, the mere existence of specialised humanitarian agencies, donor administrations such as ECHO, USAID and Danida, early warning systems, standards such as SPHERE, specialised information structures (IRIN, ReliefWeb) and coordination bodies such as OCHA, the Steering Committee for Humanitarian Response (SCHR) and the consortium of European Voluntary Organisations in Cooperation in Emergencies (VOICE) ensure some kind of basic response to most major or medium-size disasters.

The 'stakeholder commitment' argument can be exemplified by the situation in Sudan. Here, the UN, local and international NGOs, the state and rebel movements, and, to some extent, even donors are brought together in coordinated annual needs assessments, programme planning, and fundraising efforts through the UN-led Operation Lifeline Sudan and the UN Consolidated Appeal Process. Other ongoing crises such as the ones in North Korea, the Western Sahara, Congo-Kinshasa, Congo-Brazzaville, and Tajikistan do have UN and international NGO presence but in these cases the number of actors is smaller, their interactions less coordinated and they form a much weaker 'humanitarian lobby' than is the case for Sudan and Angola.

Floods in Mozambique and India: media framing is all

Sampled TV stations and newspapers provided five times more coverage to Mozambique's January 2000 floods than they had to the cyclone which struck Orissa three months previously. It seems as if the intensity of media coverage explains why Mozambique received seven times as much aid as India. This was not proportionate to the death, destruction and humanitarian need. While in Orissa at least 10,000 died and 12.6 million were affected, in Mozambique there were only 800 fatalities and some 1.5 million temporarily displaced.

Why did the Orissa emergency not grab the headlines and aid? Media access was significant. Reporters were confined for days to the Orissa state capital while affected areas were declared off-limits. In Mozambique, by contrast, transport of reporters to the disaster zone was facilitated by the authorities, aid agencies and the South African Air Force. Unhindered access led to dramatic and compelling footage. Key to the development of Mozambique as a major news story was the framing of the media coverage. Never before had viewers seen a woman give birth to a child in a treetop while helicopters hovered overhead or so many people whisked from treetops by helicopters. In Orissa, however, by the time the media was finally allowed full access to the worst-affected coastal areas, international interest had waned and dramatic footage was no longer to be had.

Africa versus the Balkans: small victory for humanitarian networking?

Examination of the media coverage of humanitarian crises in Kosovo, Sudan and Angola reveals an interesting but not surprising pattern. Kosovo won hands down. In the first quarter of 1999 there were some 5,000 articles on Kosovo while less than 450 articles on Angola and Sudan put together. Aid allocations to the Balkans tripled from 1998 to 1999. This appears to confirm the assumption that media coverage is a decisive factor in relation to the allocation of emergency assistance. But is this really the case?

While it is difficult to obtain accurate figures for the number of people in need, information from FAO/WFP crop assessments and UN CAP shows the magnitude of acute needs in the three emergencies:

  • In early 1999, around 1.5 million people were directly affected by conflict in Kosovo. Some 900,000 fled to Serbia, Macedonia or Albania while a further 600,000 were displaced within Kosovo. In July 1999, most of these people started returning to Kosovo but they returned to towns and villages which were often partly or totally destroyed.
     
  • An average of at least 2.4 million Sudanese were in absolute need of food aid during the period 1998-2001.
     
  • Using a conservative estimate, between 1 and 1.8 million people were in need of food aid in Angola in the period 1998-2001. As with Sudan, ongoing conflict prevented humanitarian access to many areas, thus preventing proper assessment of need.

 

A short dramatic war and a refugee crisis in south-eastern Europe attracted five times as much aid per needy person as the protracted wars and humanitarian crises in Angola and Sudan. It should not be forgotten, however, that Angola and Sudan received a not insignificant amount of emergency assistance in 1997-2001 ($90 to $440 million per year) despite the absence of media coverage. The widespread conviction in the aid community that the Kosovo crisis 'stole' or diverted emergency assistance from Africa to Europe (the Balkans) is thus difficult to substantiate.

In the last five years Angola and Sudan have attracted significant levels of emergency assistance, albeit far below actual needs. This somewhat surprising observation can best be explained by an influential humanitarian presence and lobby networks. A large number of UN agencies and major international NGOs have been engaged in humanitarian operations in both Angola and Sudan for more than a decade. They have well-developed fundraising and advocacy tools, enjoy direct access to donor bureaucracies and work continuously with journalists to generate a low but constant level of publicity.

North Korea, Angola and Sudan: security rules?

From 1997 to 2001 (with the exception of 1998) North Korea received more emergency assistance than Angola and Sudan. The actual scale of North Korea's humanitarian crisis is unquantifiable. Although Pyongyang prevented country-wide needs assessments, data from FAO/WFP suggests that four to seven million North Koreans needed food assistance in 1998-2001.The official explanation of the occurrence of 'freak weather' masks the reality of a near bankrupt state rocked by the collapse of subsidised trading agreements with the former communist block. With the media denied access, it was hardly surprising that there was little coverage and that the news which did emerge was based on interviews with aid workers. Neither has there been stakeholder commitment; strict conditions imposed by the North Korean authorities have limited movement of agency staff and denied advocacy opportunities.

The reason why North Korea received relatively significant humanitarian assistance can best be explained by the strong security interests of the largest donor of food aid, the US, and the shared fears of China, Japan and South Korea of the development of North Korean nuclear capability and/or an outpouring of starving refugees.

Afghanistan pre and post 11 September

Grossly insufficient levels of funding to Afghanistan throughout the 1990s can be explained by the lack of media attention combined with the absence of any real donor interests in Afghanistan. 11 September prompted a dramatic increase in assistance. In the final quarter of 2001, programmes in Afghanistan received $433m, while in the previous three quarters they had only received $232m. Security interests and the post-11 September explosion of media interest explain the growth in emergency assistance. Media coverage statistics confirm the view that it is actually the politicians, and in particular the US administration, who decide the agenda for international media attention. Correlation between the growth in the volume of emergency assistance and the American military intervention in Afghanistan is conspicuous. It could be argued that emergency assistance to Afghanistan has been an instrument for crisis management.

Interestingly, in the years preceding 11 September, EU Member States were considerably more willing than ECHO to finance humanitarian assistance to Afghanistan. While there was a three-fold increase ECHO-funding to Afghanistan from 2000 to 2001, EU member States gave eight times more. This difference could be explained by a greater inclination of national donors to act in response to emergencies that receive media attention. It is also easier for national governments to find additional money during the financial year than it is for a multilateral donor like ECHO.

Conclusion

Of the cases analysed, only Mozambique supports the oft-repeated argument that media coverage is crucial in determining the level of emergency aid allocation. It seems that the media play a crucial role in influencing aid funding decisions only when there are no vital security issues at stake. In other words, natural disasters and complex emergencies have a greater tendency to become forgotten crises when Western governments have no vested security interests in the afflicted regions. Since many disaster-prone areas - especially in Africa - are of little strategic concern to Western decision makers and since media coverage is often very limited in connection with protracted conflicts, the factor that determines levels of emergency aid will often be the degree of stakeholder commitment - the strength and persistency of the network of humanitarian organisations operating on the ground.

 

Gorm Rye Olsen is Senior Researcher at the Centre for Development Research, Copenhagen (www.cdr.dk). Email: gro@cdr.dk. Kristian Høyen is a Research Assistant at the Centre for Development Research. Email: krh@cdr.dk. Nils Carstensen is an Information Officer with Dan-Church-Aid. Email: nic@dca.dk

This is a greatly shortened version of a paper presented at a Conference on the Role of the Media, Decision Makers and Humanitarian Agencies in Copenhagen in October 2002. The full text is online at: www.forgottencrises.dk/ForgottenCrises_Conference_net.pdf.

 

Disclaimer
Opinions in FMR do not necessarily reflect the views of the Editors, the Refugee Studies Centre or the University of Oxford.
Copyright
FMR is an Open Access publication. Users are free to read, download, copy, distribute, print or link to the full texts of articles published in FMR and on the FMR website, as long as the use is for non-commercial purposes and the author and FMR are attributed. Unless otherwise indicated, all articles published in FMR in print and online, and FMR itself, are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) licence. Details at www.fmreview.org/copyright.