Fragile States and Agile Aid

I spoke at the Lowy Institute today, suggesting a few ideas for improving Australia's aid program.

Fragile States and Agile Aid: Some Ideas for the Future of Australia’s Development Assistance Program

Andrew Leigh
Federal Member for Fraser

Lowy Institute

18 May 2011

I acknowledge the traditional Indigenous owners of the lands on which we meet today.

Aid Works

Dai Manju lived in a small village in central China. Because her parents were ill and couldn’t afford the cost of sending her to school, she dropped out. When journalist Nicholas Kristof visited in 1990, she was hanging around the school hoping to pick up bits of knowledge.[1]

After publishing a front page article about Dai in the New York Times, Kristof was chuffed to receive a donation of $10,000 from a reader. He promptly sent it on the school, which spent it on improving facilities, and provided Dai with a scholarship to stay in education so long as she passed exams. After a good amount of the money had been sent, Kristof phoned the donor to thank him for the generous gift. It was only then he realised that the man had in fact only sent $100, and the slip of a bank teller’s fingers had multiplied it one-hundredfold. Informed of their error, the bank agreed to provide the difference as a donation.

After graduating as an accountant, Dai now runs her own business employing many others, and is able to send money back to improve the lives of her parents. Because many of the other girls in her class also received ‘bank error in your favour’ scholarships, the village now has a road.

Spent wisely, foreign aid can transform lives for the better. This year, Australia’s foreign aid program will build 2000 schools in Indonesia, fund a Women’s Crisis Centre in Fiji, prevent 8000 cases of blindness in East Asia, and provide clean water to 1.2 million people in Southern Africa. World Vision CEO Tim Costello claims that this year’s aid budget will save 200,000 lives.

Yet too often, aid debates become removed from the people that they are intended to help. When aid advocates ask the public for donations, they typically ask for help to solve an immediate and practical problem – feeding a child, building a well, providing schoolbooks. To the best of my knowledge, no aid campaigner ever knocked on a door and said to the resident: ‘we’d like you to consider increasing your aid contribution to 0.5 percent of your income in 2015-16, with the ambition of going to 0.7 percent in the future’.

Don’t get me wrong: the quantum of aid is important. I recognise that Australia’s aid contribution as a share of national income still places us in the bottom third of donor countries. Scaling up our aid to half a cent in the dollar will put us at the developed-country average.[2] This is eminently sensible given that 18 of our 20 nearest neighbours are developing nations, and we sell over $90 billion worth of goods to developing countries each year. Few developed nations have as much self-interest in running a good aid program as Australia. But we should never forget that the purpose of foreign aid is to alleviate disadvantage. More aid creates the potential for reducing poverty – but it will not happen automatically.

Although the input-wallahs would occasionally have us believe otherwise, the most interesting debates in foreign aid are not about the quantum of assistance.[3] So today, I want to suggest a few ideas about how to increase the effectiveness of the Australian aid program. This is particularly important in the context of the scaling up of Australia’s aid program. Over the next four years, the Australian aid budget will almost double in nominal terms. This represents a unique opportunity to ensure that our aid is spent in a way that does as much good as possible.

For an economist like me, that means identifying Australia’s comparative advantages in foreign aid, and rigorously evaluating our aid programs. Let me deal with each in turn.

What’s Our Comparative Advantage?

The principle of comparative advantage is that individuals and nations should specialise in things that we do best. If I’m a good cook but a lousy storyteller, and my wife has the reverse traits – then it might make sense for me to cook dinner, and my wife to put our kids to bed. Internationally, the principle of comparative advantage suggests that economies should ‘stick to their knitting’, rather than trying to produce everything that they wish to consume.

Comparative advantage applies to altruism too. Where there is need in the world, it makes sense for Australia to focus on the foreign aid strategies that we are comparatively best at. Since our aid will always remain a small share of global development assistance, it will do most good if it is focused.

In part our aid program is defined by our place in the world. Ours is the largest economy in a neighbourhood dominated by developing countries. We will have to remain engaged with our neighbours a wide range of development issues.  Unlike nations in Northern Europe, we do not have the luxury of being highly selective in the countries we choose to engage with.  Basic services and governance in countries like Indonesia, East Timor and Papua New Guinea rightly remain a core focus of our aid program.

But beyond basic service delivery there are still enormous needs in the developing world and limited resources to meet those needs. So our aid program needs to prioritise where it can be most effective. As I see it, Australia has three comparative advantages in foreign aid. Our first comparative advantage is that we are unusually well endowed in natural resources. This is particularly important in light of the ‘resource curse’ – the fact that developing nations who have more natural resources are more likely to be poverty-stricken dictatorships. This ‘resource curse’ arises because mineral endowments tempt despots into fighting their way into power and filching the wealth.  It’s difficult for an autocrat to steal incomes from farming, industry or services. But diamonds are a dictator’s best friend.

The curse can be seen today in the Democratic Republic of Congo, where conflict over the country’s minerals has grown particularly fierce since the mid-1990s, with bands of thugs murdering five million people, raping half a million women, and impoverishing a nation. The vast majority of Congo’s population would be better off if their country had no natural resources.

Oxford economist Paul Collier points out that if we can help developing nations to make better use of their natural resources, the resulting fiscal flows could help societies to transform themselves for the better.[4] In developed nations, oil and mineral assets generally benefit the entire population. But in most low-income countries, the opposite is true.

If developing countries can benefit from their minerals, the payoff could dwarf anything that aid might hope to deliver. Collier points out that in rich nations (where geologists have carefully surveyed the land) the typical square kilometre has subsoil assets worth US$114,000. In all probability, the same is true for the developing world. On those figures, Africa’s natural resources would be worth $3.5 trillion, more than 70 times the amount of foreign aid it receives each year. Indeed, $3.5 trillion is probably an underestimate. In recent years, the petroleum frontier has shifted south from the Middle East to West Africa. Central Africa is rich in everything from gold to coltan (used in mobile phones). Southern Africa has bountiful reserves of precious stones.

To help developing nations make better use of their natural resources, a group of ex-politicians and entrepreneurs (working with academics like Collier) have proposed a Natural Resource Charter, which they hope will be adopted by governments, businesses and NGOs.[5] The Charter aims to go beyond the ideological slanging match that has characterised natural resource use in developing nations, and offer practical ways in which governments can ensure the people get a better deal.

First, the Charter proposes that financial flows be fully transparent. Mining generates relatively few jobs, so what happens to the royalties is critical. Through the Publish What You Pay campaign and the Extractive Industry Transparency Initiative, mining companies (such as the estimated 230 Australian mining companies currently operating in Africa) are encouraged to release information about the payments that they make to governments. This makes it harder for corrupt officials and politicians to siphon it off into private bank accounts, and enables citizens to pressure governments into spending the money on much-needed infrastructure, such as hospitals, schools and roads.

Second, the Charter argues that extraction rights should be sold by auction. Once a handful of bidders participate, it becomes difficult for them to collude, and the final price is likely to reflect what the rights are actually worth. Collier uses the example of the UK, which was on the verge of negotiating a £2 billion deal to sell mobile phone spectrum when it was persuaded to try an auction instead. The auction raised £22.5 billion. His pitch to developing countries is simple: ‘if the UK Treasury can get it wrong by a factor of ten, what makes you think you’ll do better?’.

Third, the Charter suggests that developing country governments should maximise the amount of information about the country’s subsoil assets. If governments or aid donors conduct geological surveys and make them publicly available, then the people of that nation are more likely to get a fair share of their natural resources. Another way to increase information is to require that auction winners begin prospecting within a fixed period of time. If one miner strikes it lucky, this will raise the sale price when nearby parcels are auctioned off.

Fourth, the Charter requires that local peoples should be made better off by mining. Before lending to extraction projects, the World Bank requires ‘free, prior and informed consent’: a principle that the Charter argues should also extend to the way that national governments manage local consultations. Provocatively, it also suggests that where mining companies promise to conduct environmental reclamation, they might be kept to their word via the use of an escrow account.

Australia’s second comparative advantage is in dryland farming. As a nation raised on poems about dry river beds, failed crops and dying cattle, it’s easy to forget that agriculture in other parts of the world tends to be a rather more stable affair. But this has given Australian farmers more experience in good water management, selection of hardy crops and animals, and management of seed stocks. Some of this knowledge is currently being conveyed through the Australian Centre for International Agricultural Research, and our involvement in the World Food Programme and the G20’s Committee for World Food Security. Wherever possible, we should export our expertise in Australian dryland farming.

Improving agricultural productivity is critical to boosting living standards. The 1960s and 1970s saw a ‘green revolution’, in which the combination of hybrid seeds, chemical fertilizers, pesticides and herbicides massively raised agricultural output. By one estimate, the green revolution saved a billion people from starvation. Yet today, African cereal yields are 66 percent below the global average, and Africa is a net food importer.[6] The continent has one tractor per 868 hectares, compared with one tractor per 56 hectares globally. Farming also has a critical gender dimension, since 70 percent of developing country farmers are women. [7] One success story has been Malawi, which has recently doubled agricultural output through the combination of modern seeds and fertilizers. We should consider whether elements of that model can be extended to other African nations, perhaps through innovative financing models.[8]

Another implication of dryland farming is price volatility. Some climate models predict increased rainfall volatility in Africa’s Sahel, a region containing Chad and Niger, where average incomes are less than a dollar a day.[9] Closer to home, Pacific island nations are likely to lose freshwater supplies and fertile land to sea water inundations. Economist Ted Miguel argues that aid agencies should work with developing country governments to provide drought insurance to rural households. Australia’s drought assistance schemes have their critics, but given Australia’s expertise in designing such schemes for our own farmers, this is a natural area for our aid workers to be involved in.

Australia also has a role to play on the international stage. As a major exporter of agricultural products, we set up the Cairns Group of agricultural free trading nations in 1986. Today, Australia has a strong interest in arguing against policies that harm developing country farmers as well as our own. If the United States were to abolish its ethanol subsidies and the European Union were to adopt a more science-based approach to genetically modified foodstuffs, agriculture in developing countries would be much more viable.

Our third comparative advantage is operating in post-conflict environments. Where our other advantages were things we chose, this is one that was thrust upon us. From our experience in East Timor and the Solomon Islands, Australia has acquired valuable experience in post-conflict environments. Of the top five recipients of Australian foreign aid, four are countries that are often described as ‘fragile states’.

Getting intervention right in post-conflict environments involves some delicate balancing acts. One of my favourite stories is told by former Economist journalist Robert Guest:[10]

‘Somalia has no government, unless you count a “transitional” one that controls a few streets in the capital, Mogadishu, and a short stretch of coastline. The rest of the country is divided into warring fiefdoms. Warlords extract protection money from anyone who has money to extract. Clans, sub-clans, and sub-sub-clans pursue bloody vendettas against each other, often fighting over grudges that pre-date the colonial period. Few children learn to read, but practically all self-respecting young men carry submachine-guns.

‘I was at one of the country’s countless road blocks, on a sandy road outside Baidoa, a southern town of shell-blasted stone walls and sandy streets. The local warlord’s men were waving their Kalishnikovs at approaching trucks, forcing them to stop. Many of the trucks carried passengers perched atop the cargoes of logs or oil drums. The men with guns then ordered all the children under five to dismount and herded them into the shade of a nearby tree. There, they handed them over to strangers with clipboards, who squeezed open their mouths and fed each one a single drop of polio vaccine.’

Robert Guest is describing vaccination work carried out by the World Health Organisation, which decided that working with local warlords to distribute polio vaccine was the lesser of two evils.

Aid to fragile states requires considerable political nous. In his Lowy Institute paper on the Australia-led Regional Assistance Mission to Solomon Islands, Michael Fullilove argues that RAMSI succeeded in stabilising the Solomons because it was preventive, permissive, regional in nature, nationally-led, supported by the United Nations, non-sovereign, police-led, and light-touch.[11] While Fullilove is cautious about suggesting that any post-conflict operation should be a ‘blueprint’ for future missions, there are clearly experiences out of our work in East Timor and the Solomon Islands that can be used in other contexts.

Part of this is about recognising that the rise of terrorism is likely to increasingly blur the lines between military, policing and aid operations. Increasingly, in places like Afghanistan, Australia and our allies are performing roles that meld all of these (what David Kilcullen has memorably termed ‘armed social work’). This demands better planning, training and integration across government agencies than ever before. For example, literacy rates in Uruzgan province – where Australia works – are 8 percent for men and 1 percent for women. Remarkable improvements have occurred in Afghanistan, but many basic measures of wellbeing are still appallingly low.  Improving basic education is one of our most powerful tools for economic development, better governance, and stopping extremism.

How do we Know When it’s Working?

If we are to move past the simplistic focus on inputs only, it will require greater certainty about how Australia’s aid programs impact on the developing world. In some areas, we have this already. For example, we know that Australian aid has immunised 900,000 children in Papua New Guinea against measles and other diseases, supported thousands of students to study in Australia (including the current PNG Secretary for Education), and rehabilitated thousands of kilometres of roads in places like Laos and Samoa.

In other cases, the right aid strategy is less clear. Take the case of bed nets.[12] It is undisputed that malaria causes around 1 million deaths in Africa, most of them children. An insecticide-treated bed net costs around $10, lasts about five years, and halves the chance of getting malaria. Yet less than one-quarter of at-risk children sleep under a bed net.

But major figures in the aid world have disagreed over the right policy response. While Jeffrey Sachs and others have suggested that people should be given free bed nets, William Easterly and Dambisa Moyo argue that people do not value handouts, and that free bed nets are more likely to end up being used as wedding veils and fish nets. Who should we believe?

To answer the question, researchers from the University of California Los Angeles carried out a randomised trial, in which villagers in Uganda and Madagascar were randomly selected to receive bed nets that were free, or merely subsidised. Their results provide a clear answer: free bed nets have substantially higher take-up rates than nets that are merely subsidised – and free bed nets are equally likely to be used for their intended purpose.

One of the most famous randomised trials in development economics is the PROGRESA experiment, which tested whether it was possible to boost educational attendance via ‘conditional cash transfers’ – money provided to poor households on the condition that their children attended school.[13] The program was the brainchild of Santiago Levy, an economics professor who was then Mexican finance minister. To avoid the program being scrapped by the next government, Levy wanted to produce evidence so irrefutable that it would have to be continued.

He did just that. By randomising the implementation of PROGRESA across villages, the study showed that conditional cash transfers were an extremely effective way of raising education levels. When the Mexican government changed, the new government changed the name (it is now called OPORTUNIDADES), but the policy remained. Conditional cash transfer programs now exist in over thirty developing nations.

Thanks to centres such as the Abdul Latif Jameel Poverty Action Lab at MIT and Innovations for Poverty Action at Yale University, we now have hundreds of rigorous results from randomised policy trials in developing nations. Just as with new pharmaceuticals, randomised trials provide extremely strong evidence for or against particular policies. For example:

  • A randomised trial of teacher merit pay in the Indian state of Andhra Pradesh has shown that it raised student performance on both high-stakes and low-stakes tests, and found little evidence of teacher cheating.[14]


  • A randomised trial in Malawi that provided financial incentives for HIV testing showed that people who discovered they were HIV positive were more likely to engage in safe sex.[15]


  • A randomised trial in Madagascar found that simply telling parents the average income gains from another year of schooling reduced dropout rates.[16]


  • In Kenya, HIV infection rates are substantially higher among teenage girls than teenage boys, a fact that is largely because of the prevalence of ‘sugar daddies’. A randomised trial showed that a program that promoted abstinence did nothing. But a program that told young girls the basic fact that older men are more likely to be HIV-positive reduced infection rates and virtually halved teen pregnancies.[17]


  • A randomised trial of microfinance in India found that it increased the fraction of families that started a small business (by about 2 percentage points), but did not seem to empower women or shift expenditure on education and health.

Earlier this year, I moved a private members’ motion in the House of Representatives calling on all government agencies to subject more programs to rigorous randomised trials. Although randomised trials are not a perfect evaluation tool, my view is that for most purposes they are superior to other impact evaluation approaches, such as matching and before/after comparisons. Yet to the best of my knowledge, AusAID, like many Government agencies, presently does no randomised trials at all.


Looking at the challenge of development, it is easy to become pessimistic.  1.4 billion people live in poverty and two thirds of these are in our region.  Every day, 22,000 children under the age of five die from preventable or treatable conditions. Millions are living in makeshift camps, displaced from their homes by conflict and economic hardship. Teenage gangs commit acts more appalling than any Hollywood horror movie producer would dream up.

And yet, there is reason for optimism too. Technologies like micronutrient supplements and mobile phones can transform opportunities for the poorest. Organisations like the Global Fund, the Gates Foundation and the GAVI Alliance (to pick only one letter of the alphabet) combine sharp-edged analysis with serious funding and are delivering measurable results.

For Australia, some of the challenges of the developing world are challenges that we have faced in the past. Since the Eureka Stockade, our nation has been devoting attention to ensuring that mineral wealth benefits all of us. Making a living from agriculture in the world’s driest inhabited continent was a challenge for the Indigenous people as for European settlers. And in our social policies, there is an increasing recognition that improving the quality of evaluations will help us make the best use of scarce taxpayer dollars. Spent wisely, Australian aid can – literally – help to change the world.

[1] The original article was published as Nicholas Kristof, ‘In Rural China, Road to School Is All Uphill’, New York Times, 3 December 1990. For an account of the subsequent donation and bank error, see Nicholas D. Kristof and Sheryl WuDunn, Half the Sky: Turning Oppression into Opportunity for Women Worldwide, Knopf, 2009.

[2] Australian aid in 2011-12 will be 0.35 percent of GNI, and will increase it to 0.5 percent by 2015-16.

[3] The same principle applies to policy debates about school funding and national defence.

[4] Paul Collier, The Plundered Planet: Why We Must – and How We Can – Manage Nature for Global Prosperity, Oxford University Press, Oxford, 2010.

[5] See

[6] Greg Mills, Why Africa is Poor: And What Africans Can Do About It, Penguin Global, New York, 2010, p136

[7] Greg Mills, Why Africa is Poor: And What Africans Can Do About It, Penguin Global, New York, 2010, p141

[8] See for example John W McArthur, ‘An International Credit Facility To Support Commercialization of African Smallholder Staple Crop Farmers, Concept Note for GAC on Poverty and Economic Development, Millennium Promise Alliance, New York.

[9] Edward Miguel, Africa’s Turn?, MIT Press, Cambridge MA, 2009, p40

[10] Robert Guest, The Shackled Continent.: Power, Corruption, and African Lives, Smithsonian Books, Washington, 2004.

[11] Michael Fullilove, ‘The testament of Solomons: RAMSI and international state-building’, Lowy Institute Analysis, 2006.

[12] This example is drawn from Abhijit Banerjee and Esther Duflo, Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty, Public Affairs, 2011. For a host of other examples, see Dean Karlan and Jacob Appel, More Than Good Intentions: How a New Economics is Helping to Solve Global Poverty, Dutton Adult, 2011. Randomised trials also raise issues of ethics and generalisability, which I deal with in more detail in Leigh, ‘Evidence-Based Policy: Summon the Randomistas?’ (2010) in   Strengthening Evidence-based Policy in the Australian Federation, Roundtable Proceedings, Vol 1, Productivity Commission, Canberra, 215-226

[13] See Abhijit Banerjee and Esther Duflo, Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty, Public Affairs, 2011

[14] Karthik Muralidharan and Venkatesh Sundararaman, ‘Teacher Performance Pay: Experimental Evidence from India’, NBER Working Paper No. 15323, NBER, Cambridge MA, 2009.

[15] Rebecca L. Thornton, ‘The Demand for, and Impact of, Learning HIV Status’, American Economic Review, 2008, vol. 98, issue 5, pp.1829-63

[16] Trang Nguyen, ‘Information, Role Models and Perceived Returns to Education: Experimental Evidence from Madagascar’, mimeo, MIT, 2008

[17] Pascaline Dupas, ‘Do Teenagers Respond to HIV Risk Information? Evidence from a Field Experiment in Kenya’. 2011, American Economic Journal: Applied Economics, 3 (1), pp.1-36

Be the first to comment

Please check your e-mail for a link to activate your account.

Stay in touch

Subscribe to our monthly newsletter


8/1 Torrens Street, Braddon ACT 2612 | 02 6247 4396 |