Does aid work?

After decades of scepticism about development aid, the west is embracing it again. In 2001, a roadmap to the UN's ambitious millennium development goals was launched, with governments signing up to the targets of halving global poverty and reducing disease and illiteracy. Tony Blair and Gordon Brown have just launched their Commission for Africa report, arguing for a doubling of aid for Africa over the next three to five years. So how much good would result from a big increase in aid? Three new books give competing answers. Very little, says Nicolas van de Walle in Overcoming Stagnation in Aid-Dependent Countries, unless we deal with the political malaise in poor countries. Quite a lot, says Sebastian Mallaby in The World's Banker, as long as we keep NGOs away from the World Bank. An enormous amount, says Jeffrey Sachs in The End of Poverty, which makes the case for a big and immediate increase in aid.

Optimism about aid was last widespread in the late 1970s, exemplified by the Brandt report. But an even bigger wave of scepticism followed. By the mid-1980s, aid was being described as a monstrous way of transferring money from the poor in rich countries to the rich in poor countries. Aid does not work, it was said, only self-discipline and market forces can help solve developing countries' problems. It seems to have taken two decades for that orthodoxy to wear off.

The new determination of politicians to reduce poverty and disease in the poorest parts of the world could produce extraordinary results. Some aid already has. Almost $100m in aid helped to eradicate smallpox by 1977. Many hope that the new determination to fight HIV/Aids in the poorest parts of the world will bring about some similar successes.

But not all scholars of aid welcome the prospect of a doubling of aid to Africa or anywhere else. Like lottery winners, these countries could suffer a number of ill-effects from such a sudden windfall gain. It could play havoc with their exchange rates, their internal markets and their budgets, increasing the risks of corruption and mismanagement. Until now, says van de Walle, aid has been used by politicians in poor countries to prolong their tenure, rather than to improve the lot of their people.

But the sceptics perhaps overestimate the likely extent of aid increases. Aid has a habit of being more pledged than delivered. After the 2003 earthquake in Bam, Iran, donors pledged around $1bn, but by January 2005 less than $115m had been disbursed. And although the US has pledged $15bn for tackling Aids, far less than that has been delivered.

Equally, donors need to stop recouping huge slices of aid through debt repayments and contracts for their own construction companies, management consultancies and the like. Rich governments say that ensuring their own companies benefit from the aid industry sustains public support for development assistance. But requiring poor governments to use donor contractors makes every project more expensive. In Africa it has effectively reduced the value of aid by up to 30 per cent.

It is also easier to give money in aid than to give poor countries real opportunities for lasting growth. Sachs makes this point and so does the Commission for Africa, which documents how poor countries wanting to export food and commodities are sometimes blocked by trade quotas and tariffs in rich countries. And when they seek to use industrial policy to build a more robust economy—the key to dramatic economic growth and poverty reduction in east Asia—they may find that such protection has been kicked away by the new WTO trade and investment rules.

If aid does increase substantially, it will need to be better delivered to be effective. At present donors often rush in with their own pet projects, typically into a small group of pet countries. Mallaby's travels take him to some such places. Little regard is paid by donors to how many other donors are rushing into that country. Equally little regard is paid to where a donor's pet project sits within that country's own priorities. Ignoring what plans a government itself has made, donors then proceed to overwhelm the bureaucracy with paperwork and negotiations. By 2003, Tanzania was producing an average of 2,400 reports a year for external donors, and being visited by some 1,000 donor missions, each demanding the attention of senior officials. That leaves officials with little time for the business of governing the country. In the words of Mohabe Nyirabu, a Tanzanian scholar, ministers are no longer accountable to their own people because they are too busy writing reports for donors.

There is little co-ordination among donors despite the fact that governments have created institutions to foster co-operation. The World Bank and its International Development Association, the UN development programme, the World Health Organisation and the Food and Agriculture Organisation exist to combine technical expertise with the pooled resources of states, enhancing the effectiveness of aid. Yet donors mostly choose to make a small contribution to each multilateral body, while delivering the lion's share of their foreign aid bilaterally. (Britain does better than most rich countries on this, directing around 40 per cent of its aid through multilateral bodies, including the EU.)

If donors improve both the quantity and quality of aid, who is to say that aid-receiving governments will use the money well? At present rich countries rely on the IMF and World Bank to set policies and targets which poor countries must meet if they want assistance. If the IMF says a country is not in compliance, in some instances donors cease lending or giving to that country. The conditionality system is a powerful one which requires governments radically to cut or reshape their expenditure, as demanded by the IMF, to privatise state enterprises, as advised by the World Bank, and more generally to open up to global markets. All this sounds sensible. But it seldom leads to growth.

According to the IMF, between 1993 and 1997 some 46 per cent of its programmes failed, meaning a lack of compliance with conditions meant that funds were less than 50 per cent disbursed. And according to other research done by the Fund, conditionality increases the volatility and unpredictability of aid. For this reason, it has been reviewing its conditionality rules, although little has changed to date.

To be effective, aid must not create and fuel incentives for corruption. Yet this condition has never been addressed in traditional conditionality. In The World's Banker, Sebastian Mallaby recounts the day in 1996 that Jim Wolfensohn, the World Bank president, came out against corruption. He was horrified to have found that the Bank had avoided the issue for so long. Donor countries are seldom much better. They have focused on good governance, defined in terms which often do not directly address corruption.

What donors have never tried was one simple form of conditionality. Let us imagine dropping all conditions except one, the so-called single criterion. This would mean withdrawing aid when a government fails to account to its citizens for its revenue and expenditure, as often happens when leaders overstay their term of office or otherwise abuse constitutions (van de Walle proposes that aid should cease to countries in which a leader has been in power for over 12 years).

This approach would face two objections. First, the single criterion would not stop a government being incompetent and pursuing economically ruinous policies, however honestly. But this objection assumes widespread agreement as to what constitutes competence in economic policy. Many would argue that a government which runs a massive budget deficit, spends a huge proportion of its budget on a non-defensive war, increases subsidies to major sectors such as agriculture and steel before elections, and lives off foreign capital inflows is incompetent, yet economists within the Bush administration make their own case for this. There is little hard evidence in favour of any one right set of policies to achieve growth. The east Asian "development state" model requires a kind of national corporatism and state capacity that few African countries possess, and in any case is seen as too reliant on public sector intervention to be popular at the IMF or World Bank.

A more serious objection to the single criterion is that it impinges on the sovereignty of recipient governments. In theory, this is why the World Bank and IMF have long shied away from the issue of corruption and governance. Yet concerns for sovereignty never prevented those institutions from using lending to bolster the geopolitical goals of their big shareholders. And similarly, the single criterion could be used to coerce poor governments to comply with donor geopolitical goals. This is a serious issue. Faced with a possible suspension of all aid, small country governments would be vulnerable to such threats. At the very least, donors and recipients alike would need a process to prevent the abuse of aid-suspension for the wrong purposes.

Could the governing boards of the IMF and World Bank provide it? They lack credibility for the job as currently constituted. Their five wealthiest members—the US, Japan, Germany, Britain and France—each have their own appointed executive directors. Virtually all other countries are in groups which elect one representative. Some 45 African countries are represented by just two directors and wield little more than 4 per cent of voting power. The problem is not that poor countries cannot control the organisations; it is that there is no real incentive for wealthy countries even to consult poor countries. Both the IMF and World Bank have been used effectively by major shareholders—the G7 industrialised countries in particular—to impose a particular vision of what countries should do to pull their economies up by their bootstraps. The G7 has done this by exercising 47 per cent of voting power on the boards of each organisation and controlling the selection of heads. US-nominated Paul Wolfowitz will become president of the World Bank alongside the European and American in the two top positions in the IMF. And the leadership of each organisation is important because every member of staff ultimately reports to the head who can promote or demote him or her at will.

All other countries who contribute to each body have relatively little voice or influence, even though they pay the borrowing charges which keep the institutions afloat, pay the running costs, provide money for reserves, and, in the IMF, even pay the creditor countries interest on their contributions. This imbalance must change if the institutions are to play a more credible and effective role in global development. At the very least, decisions should require a majority not just of weighted votes but of all members. This would create an incentive to consult outside a core of the wealthy countries.

NGOs could have a major role to play in making the single criterion work. They have pushed the Bank and the IMF to open themselves up to public scrutiny and debate. They play an equally important role within countries. At present NGOs are attacked in Washington for being unaccountable and impinging unacceptably on the business of governments and companies. Mallaby, reflecting this view, argues that NGOs hurt the poor by stopping dams and the like being built for environmental or cultural protection reasons. But he neglects the notorious corruption that tends to surround such projects. Large electricity companies based in rich countries use their influence with their own governments or senior Bank officials to win contracts to build dams in poor countries. One result is that the Bank finds it difficult to provide impartial advice on alternatives to such contracts. Mallaby fails to mention, in the case of the dam at Bujagali, Uganda, that he cites, that the borrowing official later left office amid allegations of corruption, and that the American contractor AES imploded in the post-Enron collapse. NGO critics have not yet suggested an alternative way to hold governments and multilateral institutions to account. They also overstate the power of NGOs, neglecting the fact that they succeed only when powerful governments join cause with them.

A new aid regime and the avalanche of aid currently promised will probably not occur. But a more effective aid regime at least needs to leave governments in poor countries to make more of their own policies and mistakes. Alternatives to this approach to date seem largely to have failed.