Gillian from St. Jude asked me to comment on an article by Jeff Sachs on the role of the World Bank. Sachs is well known for his theory that the only way out of poverty for Africa is to double, triple, or quadruple aid. Gillian supports the view that such aid is necessary, indeed critical, and has often criticized my skepticism for, indeed opposition to, development aid.
In the article Sachs argues for the World Bank to worry less about corruption in Africa. He urges the new chief to instead channel more money into Africa, simply because “African governments do not have the fiscal means to invest in what’s needed, and that would be true even if Mother Teresa were running the local treasury.” His basic premise is correct, that a) Africa needs massive investments in infrastructure, and b) African governments do not have the means for many of those investments. Sachs concludes by suggesting the only player that can provide this investment is the World Bank.
While the premise is correct - in itself - it is not complete and the conclusion does not follow. But before unraveling the macro argument let us look closely at the micro examples he uses to support his point.
The Micro View: The Track Record of External Aid in Ending Poverty
The article uses four examples to justify the need for massive, external, aid. Two in particular are interesting, for they do not stand close scrutiny.
The first is the Green Revolution in India, in which Sachs claims the “the Rockefeller Foundation brought high-yield seeds, and the U.S. government shipped massive amounts of fertilizer”. Sachs does not mention that high-yielding varieties (HYV) were only one of three factors contributing to the “green revolution.” Second, as pointed out by the Cato Institute, support for HYV development accounted for only 2% of foreign aid to India in that period. So, its success does not suggest that massive aid is necessary - only that small amounts of aid can be very useful. Finally - and most important - there is sufficient disagreement amongst academics on whether the “green revolution” even occurred. As the Cato paper points out, a number of studies reveal that while wheat yields did rise substantially, “the overall rate of growth of agricultural output did not accelerate after the Green Revolution.”
The success of the second example Sachs uses - the Millenium Village Project - is even more dubious. Sachs writes that
Seasoned practitioners not held back by ideology and posturing know how rapidly results can be achieved. The philanthropy Millennium Promise, which I helped start, has raised over $100 million in private funds for Millennium Villages. This demonstrates how rural life can be improved dramatically from one season to the next.
It is interesting that Sachs’ laces his arguments with normative attacks on his readers - anyone challenging the Millennium Project, must not be a “seasoned practitioner”, nor “objective”. But to declare success now is to put the cart before the horse. That he raised $100 million for the Project does not indicate that such villages can permeanently “transform” rural life, only that he is a consumate fundraiser, to whom people will give generously. I visited the Sauri village myself, and hold little hope for its long term prospects. In an even more damning criticism of the project, Sam Rich writes
This is not to say that Sauri cannot change, or that investment in the village is wasted. But if Sauri is to become a useful model for development on a bigger scale, and not just another development expert’s white elephant, Sachs and others working on the project must acknowledge that they are still learning about Africa. Sauri is not yet a success.
The Macro View: The Problems of Massive Aid
Sachs’ overall argument for large amounts of aid also fails to convince on the macro level. Indeed, there is no better case study of the failure of Sachs’ hypothesis than India.
India has been the largest recipient of aid, ever, and between 1951 and 1992, is estimated to have received $55 billion in IDA. The Cato paper, which analyzes that entire period was published just before India had a balance of payments crises. Its conclusions are almost clairvoyant, given the timing. While the paper is painful reading for an Indian, the conclusion is unequivocal:
Except for a few cases of alleged foreign aid success– such as critical food relief when millions were on the verge of starvation in the early 1950s and again during the mid-1960s–foreign aid to India has been an unmitigated disaster. It has acted as both a catalyst and an encouragement for the politicization of the Indian economy. It has supported central planning and facilitated the growth of the public sector at the expense of the private sector and the establishment of a private-property-oriented market system. It has also encouraged corruption, rent seeking, and graft in the Indian economy. Foreign aid has been–and continues to be–predicated on an outdated and false theory of development economics that assumes that only capital and access to technology are needed for economic development.
What Sachs proposes is nothing new. Centrally planned resource allocation and development has been tried before. And it has failed.
Causality Which Way: Governance or Resources?
Sachs argues that the problem for Africa is not corruption or bad governance - something Wolfowitz focused on too much - but lack of resources. He views the challenge as an either-or, arguing that huge amounts of resources should be invested regardless of governance. But what if the problem is lack of resources and bad governance? Which should come first? I would venture, good governance.
There is good reason to believe that where governance is good, resources are not a problem. Again, India is a good example - in the post-reform period it has tripled its economic growth rate. The July 5 issue of The Economist presents another - Mali. It is a country where “Western governments and aid agencies, to say nothing of Libya, the Islamic Development Bank and the Chinese, are all flocking.” The secret of Mali is a Mr. Amadou Toumani Touré, the leader of the country, who seems to have institutionalized good governance and remains committed to reducing poverty. As a result, Mali has received USD 460 million - a massive amount compared to its public budget of USD 1.5billion.
This aid neither proves nor disprove Sachs’ theory - it is much too early for that. But it suggests that where governance is good, money will come. In Freakonomics, economists Levitt and Dubner found a similar counter-intuitive trend in election campaign finance - that it was not money that led to success in elections (the myth that money buys elections), but rather that perceived success in elections led to more money being contributed to a candidate.
Sachs holds the former view - money buys development. But Mali suggests the causation here is also inverted. It is not money that buys development, but rather the prospects of development, underlined by good governance that bring money to the table. If that is indeed true, it would conclusively weaken Sachs’ argument that the Bank should not be focussing on corruption, but on resource transfer.
Many thanks for your detailed thoughts!
You and I have common ground — that a) Africa needs massive investments in infrastructure, and b) African governments do not have the means for many of those investments.
The problem, then, is to consider ways and means for providing the needed infrastructure, without harming economic and social systems or individual livelihoods. Aid has got to be part of the solution because market forces won’t take care of it.
I think I see the beginning of signs that those who bring aid to African countries are learning from past mistakes.
One mistake in the past was to set up parallel projects, bypassing local governments, with the result that countries did not build internal capacity.
I see three examples in Tanzania where major aid is being given through the government. One is that 39% of government revenue is provided by countries such as Britain, Denmark, etc. The second is a $60million initiative from USAID to build capacity in the Tanzanian health care system. The third is a multi-country aid program of something like $200mill over several years to build capacity in the management of natural resources, especially water. Some of this aid will go to new software and hardware, and new systems, like databases. And some will go to training.
This looks like effective aid to me.
Another sign of effectiveness is that donor countries are beginning to tie aid to performance benchmarks that encourage better governance. For example, one benchmark for Danish aid to Tanzania was the enactment of new legislation aimed at countering corruption (e.g. protection for whistle blowers). When the government dragged its feet on shaping the new legislation, they forfeited a proportion of the Danish aid budget - about $2.5 million, I think.
So, I suspect that the discussion may be best framed in terms of HOW the aid is delivered, more than whether or not there should be aid.
Regarding Sachs view on resources vs corruption, I believe that he is pointing out that corruption is not the whole story, and may be not even the primary contributing factor. If corruption in an African country is about the same as corruption in other countries that are not poor, then we need to look deeper to understand the causes of the poverty. And I’m not sure he would support the stance ‘donate the aid despite the governance’. I believe that he tries hard to circumvent corruption by putting the aid as directly as possible in the hands of those who need it most.
It is certainly true, as you say, that where governance is better, donors have more confidence in giving. It is also true that economic development is stronger when there is less corruption. Corruption compromises both aid and trade - so in this discussion let’s not tie it only to aid.
More aid will work, more aid is needed. But it must be used more effectively - learning from mistakes made in the past. Even the USD460million that Mali has received is chickenfeed compared with what they need.
Thanks again for discussing this so wisely.
Gillian,
Thanks for your quick and informed response. First, I must thank you for the reference - very interesting, and challenging.
My fundamental problem with Sachs remains the same - aid delivery is subject to far too many contradictions in our imperfect world. Delivery through the public sector is inefficient and can create a massive state (see the example of India). Conversely, “direct” delivery, as Sachs wants, undermines national institutions. Finally, a recent evaluation of the Global Fund projects concludes that projects run through the government underperformed those conducted by NGOs and the private sector. So where is the balance?
As you say, aid can be part of the solution. Your examples are very promising, but in the end there is still no simple answer to the “how”. Aid is not a substitute for the right national institutions, which can be the only primary driver for development. No amount of aid will help a corrupt, badly governed country from growing. And lack of aid will not prevent a well-governed country from improving its prospects.
Dweep,
Yours is a valid point.
Talking only about resource transfer without emphasising on corruption and how efficient it will be spend is bad economics. Moreover, it is a wastage of resources.
Money and development. I do not know if we can see which is needed first; but both can lead to each other. With development, money will pour in. Also money can used (the degree will vary on the amount of corruption) to finance infrastructure building, which along with good governance can bring about development.
Again, citing corruption as the reason for not transferring funds is not warranted too.
Hi Dweep,
Here’s a new paper that finds no correlation between aid and growth.
Aid and Growth: What Does the Cross-Country Evidence
Really Show?
(forthcoming, Review of Economics and Statistics)
Raghuram G. Rajan and Arvind Subramanian
July 2007
You can find it here — http://www.iie.com/publications/papers/subramanian0707.pdf
The authors note that they saw robust findings on the importance of institutions and policies for growth, but not for aid. Maybe the new approaches that tie aid to improved institutions and policies will have an effect, not through the aid, but through the improved governance! Of course, that will depend on donors getting it right with respect to their demands re governance - the World Bank doesn’t have a good track record in that respect.