I sold my first car in Maradi, Niger - epicentre of the current famine in the country. I was particularly proud of the car as it had just managed to cross the Sahara desert, but the price I got marked me out as a future aid worker rather than businessman. Fifteen years later, there is a crisis in Maradi, and I find myself reflecting over a career spent working in famine relief, mainly in Sudan, where there are many similarities with the crisis in Niger.
In the south of Sudan, warnings of approaching famine were announced quietly by the food economy unit of Save the Children and the World Food Programme in late 1997. What could be called "intelligence chatter" had been received by Sudanese people working for these agencies. But the UN's Operation Lifeline Sudan had by that stage been running for almost 10 years, and locals had often cried wolf. So aid workers waited until they had nutritional data to back up the anecdotal reports of a famine that locals named "with a bell on its foot" because it was announcing itself from afar. At that moment the television cameras arrived and provoked a major aid effort that cost $1bn.
How can these lessons about late reaction still not have been learned seven years later? Britain's development secretary, Hilary Benn, said in one interview that "the current system doesn't work", and that "we need a better system". The reaction of aid agencies this time was to repeat that the UN in Niger had been warning of famine since November 2004 but had not received the funds it needed. Once again, they were holding up their hands and insisting that they were just cogs in the machine. Without the self-awareness needed to recognise that all parts of "the system" must take responsibility, it is by no means clear that it will change.
Reports by George Alagiah from Sudan in 1998 pointed to the lack of coordination among agencies, at which stage agencies closed ranks and blamed diversion of their inadequate aid by local warlords. In 2005, aid agencies in Niger blame aid donors for not responding to their requests for funds. Are they unaware that recipients as well as the person on the Clapham omnibus consider donors, the UN and the NGOs to be one unified system - and expect those within the system to ensure that it functions as a whole?
Let's get this straight: aid agencies did not cause the famine, so there is no need to be defensive on that count. But we have the means to avoid hunger in a relatively stable country such as Niger, and organisations cannot help but feel the spotlight of blame burning on their backs as they undertake the arduous task of picking up the pieces of a crisis that should have been nipped in the bud.
Given that aid agencies are not the ones who feed the majority of African citizens - they feed themselves - what gives these agencies the right to wade in and take control in a crisis? Is this slowness to react in Niger not proof that those who call for the dismantling of the aid structure and its replacement by equal terms of trade are right? In my experience, such big debates about corruption, neo-colonialism or dependency are the very mealy-mouthed theories that get trotted out in a vague attempt to exercise intellectual mastery over basic logistical issues or to excuse laziness and inertia.
The aid system should be held accountable when its own funding delays cause needless deaths. It would be naive to believe that there is no corruption or to ignore the bigger picture. But is it worse to withhold a food ration for fear the famine will never materialise or to give aid before people actually start dying? Far from creating a dependency, such handouts are used effectively by local families. And the financially more significant cash sent as remittances by African migrants is received and spent with imagination and economy.
Aid agencies should be accountable when they fail to develop sufficient understanding of local cultures and so ignore what is patently obvious to local people, like a famine with a bell on its foot. In my own experience, it is extremely difficult to call a famine - I watched mute and uncomprehending as one developed before my eyes in 1998. Foreigners cannot see the subtle signs that indicate when things are normal and when they are not. They do not delegate responsibility to those local people who know what is happening, fearing that a local might cry wolf or organise for aid to go to his own village.
And so the system of prediction and reaction in advance fails. In a bad year in Sudan, such as 1998 when there were fresh outbreaks of fighting and a drought, locals subsisted by sharing all they had with relatives displaced from fighting - until they themselves ran out of food and were tipped over the edge. The host families that had been masking the signs of the food shortage then actively became part of what had suddenly become a famine.
Does aid have a role when it can get things so appallingly wrong? In these failures, perhaps there is an opportunity to question the right of the charity business to have a monopoly on "calling famines" or to claim moral superiority over people who quite competently feed their families and share unquestioningly with their extended families in normal years. They suffer because they are poor, not because they can't look after themselves and their families.
· Simon Harragin is a freelance anthropologist with experience of famine relief in Sudan
• This article was amended on 14 December 2015 to correct the spelling of Simon Harragin's name.
