The United Nations announces a famine and that 12.4 million people are threatened by drought in the Horn of Africa. Radio and television repeatedly broadcast an appeal for donations to UNICEF, brandishing disturbing figures. Oxfam accuses the French government of being tight-fisted.  Rony Brauman questions the use of such language of catastrophe and of such numbers being quoted in the media like rabbits being pulled out of a hat.

Olivier Falhun: What is your analysis of what is happening today in the Horn of Africa?

Rony Brauman : Despite the lack of images and of verified and substantiated information, the mere fact that every day between 1,000 and 3,000 people cross the Kenyan or Ethiopian border demonstrates the seriousness of the situation: people do not take the decision to flee their homes lightly. In other words, there is no doubt that the situation in Somalia is serious. But in saying this, I remain rather vague, because I cannot tell you what is really going on. Are the people who leave fleeing a war zone or persecution? Or is it that there is no more food available? Both reasons combined, probably. Despite the numerous analyses, accounts full of figures, and UN and NGO reports, we still don't know.

Given the number of countries affected, it is likely that this situation is more related to lack of food than to violence ...

A survey published by Fews Net reveals that southern Somalia is the most affected, the north is enjoying relative abundance and the centre is closer to an imbalance. But agriculturally speaking, the south is the most modern region. It is a valley that lies between two rivers, the Juba and Shabelle, and is traditionally an exporter of food. If subsistence if threatened here, it is likely that elements linked to political instability are also a factor. There must be a relationship between the effects of the drought and political instability, if only because in this region of southern Somalia the weather conditions are normally offset by irrigation systems, systems which have not worked this time. This situation is therefore the consequence of a lack of rain coupled with political failure, even if it is difficult at this point for us to know exactly what happened. The situation - drought and conflict - is comparable in Ogaden [in Ethiopia], where the movements of NGOs are subject to severe restrictions and where it seems impossible to make a diagnosis of the food situation.

Should we then question the magnitude and extent of the crisis affecting the whole of the Horn of Africa?

This figure of 12.5 million people spread across half a dozen countries has no significance other than as an alarm signal from the UN to alert us to an extreme crisis. But I think this is a bad strategy. First, by using six-figure numbers and counting victims by the million, you end up demonstrating your insignificance, and the weight of the horror crushes the very people you are trying to mobilise. Moreover, if journalists were not stricken with amnesia, those who advance such figures would be exposing themselves to ridicule. Two or three years ago, Oxfam spoke of 30 million people threatened by starvation in the Horn of Africa. In Burma, there were rumours of a million and half people in danger of imminent death. Not to mention the two billion people at risk of flu! Year after year, millions of "virtual" deaths appear on the front pages and are quickly succeeded by others. Then they are almost immediately forgotten, which emphasises the insignificance of these announcements. We should move beyond the numbers and focus on where the food crisis seems to be the most critical: parts of southern Somalia, where there are pockets of famine. Note also that the uncertainties are due to the lack of access to affected regions, which brings us back to the responsibilities of the Somali political leaders.

How do you reply to those who worry about a timebomb, who fear that current food insecurity might lead to a future large-scale famine?

I would tell them that they do not know, that famines are difficult to predict, and that to talk about a potential famine is an abuse of language. Some are now talking about a famine spreading across the whole of East Africa, like an epidemic. It does not happen like that. It is never entire regions that are affected, only localised areas. Moreover, Kenya, Uganda, and South Sudan differ in many ways. Grouping them together in a "Horn of Africa" makes no sense in this context. Finally, a famine is immediately visible. We can see it straightaway, when a significant number of adults are starved to their bones, and this is not the case in the areas we have been able to access. My aim is not to get caught up in an irrelevant semantic debate, but we must remember that the words employed to describe a situation will determine the type of response. To use medical terminology, it's about accurate diagnosis and appropriate treatment.

What do you think in this case of Oxfam's pronouncements against the French government, who they accuse of selfishness in a crisis of this magnitude?

From a marketing point of view, it was well done, as the French contribution has jumped from 10 to 30 million euros. But, contrary to what the current campaigns would like us to think, it is not enough to accumulate millions of dollars to help the Somali people. So this is only a short-term victory: once they have collected the money, it will be difficult both to provide a precise account of where the money was spent and to guarantee the provision of relief.

Aren't the NGOs also guilty of spreading this idea that money miraculously results in relief, although this probably does not reflect reality?

I think the idea is in tune with the current belief in the omnipotence of money and technology - a belief that NGOs also repeat and recycle in their own field. The war in Libya is another example. I use the term "belief" because I relate this approach to a kind of religious behaviour, a kind that sits very happily with more material interests. Remember the controversy over donations after the 2004 tsunami. The fact that MSF dared challenge the urgency of the situation and refuse donations resulted in unusually violent reactions that can only be explained by some kind of believer's passion, combined of course with the desire to fill bank accounts. What is happening in Somalia reminds us that you can't solve famine by air-dropping food to an isolated and distressed population. Yet this "model of providence" is the one that is being disseminated: you just need to raise funds, convert them into food, and distribute it to all those who need it. Unfortunately, things are not as simple ...

Given the increasing media pressure and the extremely alarmist reports, I would not be surprised if military escorts will soon be suggested for aid convoys. It would be unwise, but it wouldn't be the first time something unwise was done. Unless we want to repeat the huge mistakes made by the US-led Operation Restore Hope in 1992, it is obvious that we must organise assistance with the local authorities and local intermediaries, whom we should neither romanticise nor demonise. Without them, or against them, nothing will be possible. This is one of the few certainties of the situation. It also means that this is their responsibility too, not just foreign players.