The first thing that you notice when you step into a feeding centre in Africa is the stench. It is the smell of the effort to clutch onto life, fetid, desperate. It is difficult to conceive of anyone being able to put up with it for long.

But, when you work in such a centre, you learn you can put up with it. It becomes a part of the environment. You learn to see that battle for survival as something proud, that your work supports, and not something pathetic, that you pity. After a while in such a place, in some famine-hit part of rural Africa, say, you come to find the successes more exhilarating than the failures are depressing. Seeing children leave on their feet (or at least happily strapped to their mothers’ backs) makes seeing others leaving for the burial ground more tolerable. Success is measured simply, in statistics: mortality, weight gain, malnutrition rates. Good stats make the work worthwhile.

Feeding centres have been at the core of the institutional response to famine for decades. They have given us our iconic images of white aid workers holding swollen-bellied black children, under plastic tarpaulin in some God-forsaken plot of parched land somewhere. Based on the hospital model, they make sense to us. They appeal to our instincts for centralised organisation and distribution. They echo deeply-embedded ideas of discipline and order. They allow aid workers to run and oversee aid programmes efficiently and systematically.

So it’s easy to forget, especially when you work in one, how unnatural they are. They bring sick people closer together. They require mothers to leave the rest of their families for weeks at a time, in order to care for one child who is sick. They prevent mothers (almost always mothers) from doing work on their smallholding. They reduce people who have been fighting for their survival to the status of “beneficiaries” of foreign charity, often administered by people who look suspiciously like the former colonials, except in t-shirts and sandals.

No matter. They save lives. In a crisis, that’s what counts. But, around 2001, an alternative emerged. A young, dreadlocked doctor named Steve Collins visited feeding centres in central Angola, where I was then a t-shirted and sandaled young aid worker, and talked about a curious idea he had: that the systems of emergency feeding could be combined with the well-established ethos of community development. He looked at our centres, helped improve them, and then wrote a report suggesting that what we should really be doing was teaching mothers how to treat their seriously-ill children in the community.

It was controversial. Sending things out into a community – whether information or goods – means forsaking control. Some situations – such as conflict – are too unstable to permit it. Getting it wrong would risk children’s lives. But Collins, backed by Concern, slowly developed the strategy, testing it, gathering scientifically measured positive results, and testing it some more.

Last year, it was backed by the United Nations. Two weeks ago, at an Irish Aid meeting in Nairobi, Kenya announced it was rolling out the strategy, now known as “community-based management of severe acute malnutrition” across the country. Malawi has already made it a cornerstone of its public health policy.

The strategy is one of the elements in the recent report of the Hunger Task Force, a group of international experts asked by Irish Aid to recommend how aid money can best fight global hunger. The report’s key recommendations have the simplicity that comes with ideas that seem, in retrospect, obvious. Focus on smallholder agriculture, it says; improving the access of poor people to the land will increase productivity. And promote breastfeeding, and school meals, amongst other ways of improving the nutrition of mothers and children.

Against a bleak background of rising food prices and increasing hunger, the report notes intriguing successes: Brazil has given cash to small farmers; Malawi has given them fertilisers (which the country could barely afford). In each case, incomes and productivity have risen, and hunger levels fallen.

There is an idea that runs through these successes, and lies implicit in the task force report. It is that even the poorest people must have a say over how they are helped, and freedom in how to use the help they are given.

It may seem, sometimes, like the business of aid is blighted by task forces and special commissions, high-level panels and international quangos. But methods change, and lessons are learned, and somebody needs to roll them out. The idea behind the Hunger Task Force was to position Ireland as a world leader in developing solutions to global hunger. Our experience of colonisation and famine makes us a natural lobbyist for those countries for whom such events are recent history, or current affairs. And it may help us give those who are struck by crises our empathy, and not just our sympathy. That, combined with good ideas, and political momentum, could be a potent force for change.

Published in the Sunday Times, ‘Think Tank’ column, November 16 2008.