杏吧原创

Humanitarian aid was distributed in Haiti after an earthquake hit in 2010 (Image: Logan Abassi/MINUSTAH/Getty)

We need to consider poverty as a complex situation if aid is to save lives in the 21st century, says Ben Ramalingam in Aid on the Edge of Chaos

鈥淓XCUSE me friends I must catch my jet, I鈥檓 off to join the Development Set.鈥� There can be few people working in international development who don鈥檛 know this about aid workers who discuss famine over steaks, spout jargon at endless meetings 鈥� and ultimately make little difference to the poor.

But in many ways the image is unfair. UN, government and private agencies trying to help poorer countries do indeed constitute an industry worth $50 billion a year. But in my experience, they are generally a hard-working, high-minded lot who really do want to help. And science is never more inspiring than when it is breeding better cassava, stopping malaria, spotting locust swarms from space, coordinating disaster relief or otherwise working to improve the lives of impoverished people.

And yet those impoverished people remain, stubbornly, impoverished. The exceptions, notably in China and India, owe their newfound prosperity to trade, not aid. Some propose, only half jokingly, that we should use that aid money to hand $1 a week to each of the 鈥渂ottom billion鈥� poorest people. But poor societies aren鈥檛 machines where you can pour money or technology in one end, and development comes out the other. Decades of futile projects stand as witness.

聯Poor societies aren鈥檛 machines where you pour money in and development comes out聰

In his recent book The Great Escape: Health, wealth, and the origins of inequality, economist Angus Deaton wrote: 鈥淲e often have such a poor understanding of what they need or want, or of how their societies work, that our clumsy attempts to help on our terms do more harm than good.鈥� He thinks we should abandon development aid beyond medical and agricultural R&D.

Development researcher Ben Ramalingam would agree with all but that last bit. In Aid on the Edge of Chaos, he explains his vision of how foreign aid could be fixed. It鈥檚 all down to that fashionable buzzword: complexity.

Ramalingam knows a lot about the development business. His detailed analysis might be too exhaustive for some, as he shows how aid can be more about reinforcing the power and preconceptions of the aid agency than ending poverty. But he also argues that even at its best, aid is a child of 19th-century science, with reductionist solutions for simple problems. Sometimes this works, but more often it doesn鈥檛. Aid doesn鈥檛 have the expected effect.

聯Even at its best, aid is a child of the 19th-century, with reductionist solutions for simple problems聰

Machine solutions

Thus the grain marked 鈥済ift of the American people鈥� put African farmers out of business, making everyone poorer. The $50 million meant to build houses in Afghanistan went mostly to intermediaries, and the lumber that did arrive was too big and ended up as firewood. The goats the World Bank gave poor Brazilian farmers were abandoned as soon as the bank鈥檚 added money and perks ended: they were more work than the farmers felt was warranted.

The mistake in the last example, says Ramalingam, was to regard the farmers as a linear, machine-like system, whose poverty could be sorted with a simple 鈥済oats in the machine鈥� fix. (He makes great puns.) When an input doesn鈥檛 have the expected output, he says, you are probably faced with a complex system.

Ramalingam knows enough about complex systems that his book would probably be worth reading as a non-mathematical primer on complexity theory 鈥� the science of non-linear, interlocking systems. Some of his best examples come from the way it has already been applied to banking and finance.

When it comes to using complexity theory to help the poor, he advises working at the 鈥渆dge of chaos鈥�, where the system is poised between too much rigidity and too much risk. Leading complexity theorist Stuart Kauffman calls this 鈥渁 compromise between structure and surprise鈥�.

It all makes perfect, if qualitative, sense. But while it seems plausible that oversimplifying poverty has caused much aid to fail, is a paradigm switch from simple to complex systems the answer? Has such a switch ever rescued a foundering aid operation?

Ramalingam鈥檚 closest example is in Bali. The traditional system used by rice farmers to allocate irrigation water adapted flexibly to complex situations. Then in the 1980s an Asian Development Bank project replaced this with modern high-yield rice, and pesticides and fertilisers were used without the traditional controls. Continuous planting allowed pests to multiply. Rice yields collapsed. Happily, the traditional system was restored.

This tells us we need to understand what locals are doing and not assume arrogantly that technology which worked elsewhere will work identically everywhere. But is it science?

Furthermore, if what the locals do is already optimally adapted, why do they need aid? Traditional solutions can certainly falter under modern conditions 鈥� bigger populations are a good example. But where classic aid ignored tradition, complexity thinking uses it as a starting point to try improvements. Ramalingam doesn鈥檛 say, but I bet the Balinese are still using some of those rice varieties and fertilisers, just with traditional irrigation controls. Maybe such combinations are the edge of chaos.

One way actual complexity science, and not just respect for the complexity of a system, has helped is through the use of networks to reveal fragilities and centres of control in social systems. Ramalingam cites water use in Ghana, humanitarian relief in Mozambique and Haiti, food aid in East Timor, and farm markets in Uganda. Network analysis has also shed light on bird flu in Vietnam.

An infant science

The Health Foundation, a British think tank, recently evaluated a similar movement to use . The think tank likes it because it challenges assumptions, focuses on relationships rather than simple cause and effect models, and provides a more complete picture of problems. But it complained that the theory isn鈥檛 well defined or tested against alternatives. That seems a fair assessment of its application to development, too.

But complexity is an infant science, warns Ramalingam. Using it is more like setting up a process to experiment with solutions than imposing a defined remedy. If some aid workers are now doing that, maybe it is because of a paradigm shift. Or maybe it鈥檚 just a realisation that simplistic, externally imposed solutions don鈥檛 work as well as partnering with poor people to understand knotted problems.

But the science still matters. There are plenty of signs . If invoking complexity science helps them catch on, good. If one day that science evolves more tested, defined ways to help end poverty, even better. Meanwhile, this book explains an important global activity few outsiders understand, and important scientific ideas that might yet turn it around.

鈥淭he poor ye shall always have with you鈥� goes the cynical last line of The Development Set poem. Maybe not. That鈥檚 still a goal at least worth pursuing.

This article appeared in print under the headline 鈥淥ut of chaos into complexity鈥�