Print
Share
Adrian Wood’s FT Proposal to Cap Aid in Africa Sets off Lively Debate — Here’s What I and Others Think
September 8, 2008
Adrian Wood, a professor of international development at Oxford, proposed in an article in the Financial Times last week (free registration required) that donors as a group limit their aid flows to aid-dependent countries to no more than 50 percent of the tax revenue a country raises from its own citizens (excluding oil and other mineral revenue). He argues that too much aid undermines a government’s accountability to its own citizens — a point made by Moss and Subramanian, Brautigam and Knack (Subscription Required), and African tax commissioners, too.
Prof. Wood sent his Op-Ed to some of the world’s leading scholars of aid and aid effectiveness, sparking a lively e-mail conversation, that has continued for nearly a week. With the authors’ permission, I’ve excerpted some of the more interesting responses below.
Among others, Michael Lipton liked the idea, and reiterated the dangers of donors financing “recurrent costs” — something the Scandinavians and the Africa Commissioners (the Blair Commission) have endorsed. But….
Jeffrey Sachs objects to any new formulas to cap aid:
What a time to be arguing for capping aid based on some new and arbitrary limits, when the donors are flagrantly violating every promise that they’ve made to increase aid. And targets like 10 percent of GNP are simply meaningless when we’re dealing with countries at $200 per capita, disease pandemics, no electricity, roads, ports, safe water. That’s a $20 ceiling. Great. There are practical things that can be done to save lives, build infrastructure, develop agriculture, adapt to climate change, put children in school, and more. Aid levels should be based on rigorously assessed needs to achieve given objectives, most importantly the Millennium Development Goals, and should be delivered in a systematic manner, based on milestones, audits, monitoring, and evaluation.
Gus Ranis, while sharing Wood’s concerns about the potentially perverse effects of too much aid, was skeptical about the value of rigid rules:
I agree with the dangers to governance that potentially accompany foreign aid or natural resource related flows of capital. However, whether this danger can be avoided depends on how the volume of aid and its purposes are negotiated, hopefully by letting the recipient take more of the initiative and determine its own “self-conditionalities.” I am also skeptical about any across the board rigid rules which make it impossible to reflect different countries’ very different problems, opportunities and tax capacities.
Here’s what I think.
First, donors do need to tie their hands with simple and clear rules — to overcome their own coordination and so-called “alignment” (with recipient country programs) pathologies. They have been trying for at least a decade, but so far without any notable progress — as is clear from their own reports to themselves and their recipient country “partners” (see the new OECD-DAC report prepared for the High-Level Forum on Aid Effectiveness just held in Accra).
So it is hard to argue with the simplicity and clarity of the proposal. (Two weeks ago we suggested six ideas donors might have embraced to lock themselves into better behavior at Accra. Only on transparency did something clear emerge however — and six ideas may be too many!) But the real problem is that Adrian Wood’s simple rule won’t work because a proposal to be applied to the collective of donors doesn’t make any one donor accountable to anyone. Oops: another accountability problem. How ironic.
Second, Wood’s proposal might well create a perverse incentive for aid-dependent governments to resort even more to trade and other indirect taxes than they already do. For why that’s bad for the incipient middle class, for small business, and for governance itself, see my working paper Do No Harm: Aid, Weak Institutions, and the Missing Middle in Africa especially p. 16.
Third, some recurrent costs are in fact “investment,” including the salaries of teachers who build the human capital of the next generation. The problem is not that donors might reasonably cover some of the costs of salaries in very poor countries. The problem is that under the existing aid system they cannot do so in ways that are stable and predictable and they cannot manage to get out of the way with their own ideas about the key inputs. They cannot, as a group, really cede ownership so they constantly risk undermining local institutions.
I agree with Jeff Sachs that more money could be spent well. And I agree with those who fear that there can be too much of a good thing. What’s needed is an approach to aid that helps, indeed forces donors, to shift accountability of recipient governments away from donors and back to citizens — allowing for the feedback governments need from their own taxpayers. We suggest one way to try that — called Cash on Delivery Aid. You can read about that here, and, soon, in our forthcoming publication: Cash on Delivery: Paying for Progress with Foreign Aid.
The debate continues, with e-mails flying fast and furious this morning. With the authors’ permission, we will post some of these comments below. I invite you to join the fray.
9 Responses to “Adrian Wood’s FT Proposal to Cap Aid in Africa Sets off Lively Debate — Here’s What I and Others Think”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
September 9th, 2008 at 11:41 am
I much enjoyed your piece and thought it was very well argued. I’ve often said (though probably not written) that, except in post-war/emergency conditions and briefly, the rule should be that aid should not exceed public-sector investment. If it does, the donors pay the current salary of civil servants. As aid increases above that cut-off, governments increasingly become accountable to donors rather than electorates/taxpayers. (Net borrowing clouds or complicates the rule a bit, but that can be handled.)
If the rule is right, the ‘Scandinavian wisdom’ that donors should fund much of the current costs of their own programmes, not just the capital costs, is exactly wrong – as is some of the Africa Commission report (other problems with that are neglect of both agriculture and population growth, i.e. it’s Hamlet without the Prince and Claudius).
September 9th, 2008 at 11:43 am
My take is that this is an elegant but not very practical idea. The “correct” upper limit of aid as a proportion of the economy is no exact science. An aid dependence mentailty is definitely bad for you, but this is more a matter of approach than statistics. Nepal over the years has exhibited a number of dependence features (eg reluctance to “say no” to donors) which Uganda, with a much higher aid/GNI ratio has usually avoided.
With Africa growing rapidly and country programmable aid growing and expected to grow very slowly (see the DAC publication on forward aid flows of a few months back), aid dependence seems unlikely to grow significantly, other than in a few special cases.
September 9th, 2008 at 11:44 am
I tend to agree with Jeffrey Sachs, and also Gus Ranis, in this discussion. Concern with rigid rules is very relevant, and it is certainly not “premature to worry about that sort of issue.” The debate on “absorptive capacity” for aid has gone on for a long time, with a literature that dates back to the 1960s. I think that focusing on rules is a mistake. The major issues come down to:
(1) Aid modalities,
(2) Productivity of the aid
(3) Resulting changes in demand for goods and factor services, particularly non-traded goods and services, given that the aid is funded by increased supply of foreign exchange.
If the aid is productive, particularly in the traded goods sectors, then there need be no Dutch disease problem. If the aid strains governance, then the issue really concerns modalities. In this case, there is no simple rule and I think that Adrian’s proposal leads the debate into sterile ground. The answers will be country and institution specific, and a discussion of simple rules leads us astray.
September 9th, 2008 at 11:45 am
May I add an ODI voice to the endorsements of your OpEd (with a footnote of congratulation to Moore and Brautigam for the recognition they have been getting through you and elsewhere for the work they have done in opening up the taxation/aid/governance topic)?
Raising this issue now seems to me exactly the right thing to do. As Jeffrey Sachs’ comment reminds us, nearly all the public debate currently centres on a very simple story-line linking unmet donor aid targets and unmet human needs. There is almost zero recognition that the relationship between raising money and meeting needs is in any way complicated. If it is a diversion to generate some public and political awareness of aid absorption problems and perverse effects, a diversion is what we need!
Gustav Ranis’ concerns about rigid rules will be widely shared, but it would be premature to worry about that sort of issue. As your article makes clear, the collective action problems involved in getting your proposal anywhere near being implemented are pretty daunting. In the meantime a simple target number serves a useful educational purpose.
But the main thing I want to say is that the case for a new approach would be more telling if it were double-barrelled — combining both of the main arguments for being concerned about excessive aid dependence: 1) the governance/revenue one, on which you concentrate in the OpEd; and 2) the macroeconomic absorption one, highlighted a couple of years ago by country case studies by Killick and Foster at ODI and others at the IMF suggesting that several African countries are already at their absorption ceiling.
Michael Lipton alludes to the latter type of argument when he mentions his rule-of thumb that would peg aid volumes to levels of public-sector investment, rather than to taxation. Michael seems to imply that this is all the same point, but I am not sure that it is, except in the sense that it leads to a broadly similar policy conclusion. The reason for worrying about the use of aid to pay for teachers’ salaries is not about the impact on governance and domestic accountability. That impact is surely very much the same for investment and recurrent costs. The reason has to do with the combined effects of aid-funded public spending and the accompanying central-bank sterilisation measures on credit availability for the domestic private sector, on the exchange rate and so on — these effects being potentially less serious for investment expenditure, which may be directed in ways that offset Dutch Disease, crowd-in private investment, etc.
This is not at all my area of expertise, but it seems to me that there ought to be scope for using some of the simpler macroeconomic arguments to buttress the governance-based arguments: an aid-to-tax ratio above a certain level will not only further weaken governance, but will also likely be spent in ways that crowd out the domestic private sector, and thus damage development prospects directly. Shooting at the conventional wisdom with both barrels would surely reach more people more effectively than shooting with just the one.
September 9th, 2008 at 11:47 am
What exactly are we disagreeing about?
Jeffrey Sachs (and Sherman) are surely right that (a) ‘The efficacy of aid is determined by the directions of use and the modalities of application. When invested in Africa in productive agriculture, health, education, and infrastructure (roads, power, water and sanitation, ports) the returns can be enormously high, as has been repeatedly proved. The modalities exist for scaling up, as long as aid is delivered within a serious management framework – with quantification, timelines and milestones, predictable financing, audits, monitoring and evaluation, and local accountability and participation’.
(b) ‘Malaria control, aids, TB, schooling, agricultural inputs, water, contraceptives, fertilizer, seed multiplication, feeder roads [are among s]pecific, urgent, working and effective aid programs .. starved for cash .. Specific, achievable investments in fertilizers, improved seeds, bed nets, rapid diagnostic kits, school-meal and nutrition programs, de-worming, roads, power, and much more are waiting for funding long promised but not delivered. The idea that such programs are generally limited by “aid absorption problems and perverse effects” is way off base’.
But Adrian Wood (and others, including me) are surely also right that
(a) some forms of aid are fully fungible either by design (programme aid) or by outcome,
(b) aid sometimes comprises a ‘large’ part of GDP (>15%?),
(c) so several governments get resources, especially fungible ones, from – and therefore become accountable to – donors much more than citizens;
(d) if (b) and (c) extend beyond postwar/emergency situations, and go on for many years, this corrodes (or stops) development and participatory politics.
Yes, ‘The efficacy of aid is determined by the directions of use and the modalities of application’, but who is sovereign? I would much sooner have Jeffrey decide the structure of aid than Joseph Kabila, but isn’t that ultimately up to the people of the DRC? Given actual decision structures, wouldn’t matters be made much worse if Kabila or Mugabe or … were to join the ‘club’ receiving massive aid that is, in effect, NOT substantially going to Jeffrey’s correct priorities?
One problem is that the shift from project aid to programme aid (via SAPs etc) went much too far. The shift, amazingly when one thinks about it, was sold on the prospectus that it reduced the problems of fungibility and ownership. It does the opposite, while making accountability much harder.
September 9th, 2008 at 11:48 am
Thanks to all of you who have contributed to the correspondence prompted by my article (and apologies to those others who feel that it has merely clogged up their in-boxes). There is more correspondence, for any who want more, in the FT forum at http://blogs.ft.com/wolfforum/ In response to Michael’s question, I see three areas of agreement and four areas of disagreement.
The three areas of agreement are (a) that aid can be enormously valuable to people in poor countries, (b) that donors should provide more aid than at present to many countries, and (c) that the value of aid to its recipients depends greatly on how it is delivered.
The first area of disagreement concerns the optimal mode of delivery, on which there is a deep division between those who favour the country-led model of the Paris and Accra declarations and those (including Jeff and
Michael) who favour a more project-oriented approach. But this disagreement was not the subject of my article.
The second area of disagreement is about whether any sort of upper limit should be imposed on aid to particular poor countries (as distinct from phasing out aid as countries cease to be poor, which is not contentious – for example, Jeff uses the phrase ‘time-bound’). The third area of disagreement is about whether any sort of an upper limit could, as a practical matter, be imposed by donors. The fourth area of disagreement is about the appropriate form of an upper limit.
My view in the second area is, as my article stated, on balance ‘yes, there should be an upper limit’ (though a minority of me disagrees with myself, on the grounds so well put by Jeff). This is less because of the macro problems that David mentions, which I think can be managed adequately (on this, I agree with Sherman), than because of the governance problems on which my article focused. And alas, on the governance front, though Jeff is right to note the tiny absolute numbers, one cannot get away from percentages of GDP (or more precisely percentages of government revenue and spending).
My view in the third area is ‘I don’t know, but it does not seem utterly impossible’. My doubts are mainly about lack of donor co-ordination, which, as Deborah points out, has been aggravated by the arrival of China. I am less doubtful than Bill Easterly (in the FT forum) about donor motivation, and I think that the formidable practical problems of defining and measuring aid and taxes that Simon mentions are soluble.
In the fourth area, I don’t much like my simple 50% rule, and on that I agree with Gus and Sherman (and with Paul Isenman, in a private message). Such a one-size-fits-all rule would clearly be inferior to a policy based on a country-specific balancing of the marginal benefits and costs of aid, and if upper limits on aid were imposed, I hope (and
expect) that they would be of this more sophisticated kind. But I wrote the article because I agree with David that ‘a simple target serves a useful educational purpose’ (and so I disagree with Sherman that it ‘leads us astray’). ‘Educational’ is not quite the word I would have
chosen: the usefulness of an over-simplified proposal is that it can stimulate, and provide a clearer focus for, a debate.
Some of you, such as Jeff, argue that this debate is entirely inappropriate (and for anyone who says ‘no’ in the second areas of disagreement, the third and fourth areas are simply irrelevant). Others of you, like me, feel that there is something here that needs to be debated. I truly don’t know what the conclusion of this debate will be, or even should be, but I hope it will continue in various shapes and forms.
September 9th, 2008 at 11:49 am
One thing we do disagree about is your assertion that a return to the old chimera of project aid ensures local ownership and accountability, in contrast to program aid. The key is whether donors are truly capable of letting recipients take the initiative, telling us which of the several possible productive purposes cited by Jeffrey the aid -and complementary local efforts – should be addressed to in terms of their own assessment of priorities and capabilities. Donors would, of course, respond,negotiate, and insist on “self-conditionalities” precedent and credible compliance. Program aid, even if badly misapplied in the SAP past, remains the potentially best, most flexible and attractive instrument for enhancing aid effectiveness.
September 9th, 2008 at 12:27 pm
Coming from an oil cursed country (Venezuela) and knowing too much about having to live with governments that are made wealthy independently from their citizens, I absolutely agree with the general concept that Adrian Wood expresses, though I cannot understand why he makes a specific exclusion in the calculations of the revenues from oil and minerals.
If we are to have sustainable democracies, where governments are accountable to their citizen, then a rule that caps all the revenue that the State obtains, whether from aid, oil or whatever not paid directly to it through non-coercive means by their own citizens, to a certain percentage of GDP, for example 5 percent, and otherwise oblige that all aid (and hopefully oil revenues too) is distributed directly to the citizens would seem like a much better alternative.
Friends, you should not impose a limit on how much help donors want to give, not with the many needs at hand, nor at how much a citizen could receive, again not with the many needs at hand, and I do not know of any such limit that has helped a developed country to develop, but you sure could help to impose limits on how to avoid concentrate the aid given and received in too few gubernatorial hands. Please go ahead!
September 15th, 2008 at 10:37 am
Adrian Wood gives some good reasons why aid donors might think about capping the amount of aid given to individual governments. Bill Easterly is fully sympathetic with the objective, but argues that it will not happen. Why? Because it did not happen in the past; and because it is easy to flourish a perfectly plausible political economy explanation: the main interest groups – aid agencies and recipient governments – just don’t want it to happen. Adrian Wood says he is more optimistic about the possibility of change. So am I.
The spate of recent trenchant critiques of aid, including Bill Easterly’s own book White Man’s Burden, are having consequences. Perhaps the most damaging evidence comes from a number of recent pieces of research revealing just how unpredictable aid inflows are from the perspective of the recipient government. For that reason alone, the real value of aid is significantly less than the figures suggest. (See for example
http://siteresources.worldbank.....le_Aid.pdf). Perhaps more important, younger African professionals are fed up with aid, aid donors, and the dependency that this inevitably implies. Those among them who staff ministries of finance and revenue authorities do not have to look far for inspiration and practical examples of how to improve Africa’s overall mediocre record of official revenue raising. Since the end of apartheid, the South African Revenue Service has performed extremely well in raising additional revenue, in conveying to the population the purpose of their paying taxes, and in establishing a reputation as one of the most trusted public agencies in the country. The Rwanda Revenue Authority, led by a much respected young professional, Mary Baine, is setting records for re-establishing an effective and efficient taxation system.
The depth of this African professional commitment was conveyed very clearly at a meeting of over 100 delegates organised by a number of African revenue agencies and the OECD in Pretoria just two weeks ago. The communiqué from that meeting, available on http://www.sars.gov.za/home.asp?pid=12840, convincingly conveys the spirit of the event. The delegates, representing 29 African countries and a number of aid donors and international organisations, exhibited a surprising unanimity, in statement after statement, speech after speech, around three main messages. The first is that taxation should be seen not just as a technical or fiscal issue, but as part of the state-building process. In other words, the kinds of arguments sketched out by Adrian Wood about the beneficial political effects of governments’ fiscal dependence on domestic taxpayers were seen to have real policy traction at home. The second message is that domestic financial resources are better than aid. And the third is that raising these domestic financial resources is very much an African responsibility. A group of African tax commissioners are already mapping out the next steps.
Is it reasonable to expect that all these good intentions will lead to action? No. Is it reasonable to expect that most African countries can replace aid with domestic taxes within the next decade? Again, no. But attitudes definitely are changing. It does not matter that much whether aid donors can or will adopt Adrian Wood’s proposal to cap aid. It is more important that they do their best to support local initiatives to wean Africa off aid in the foreseeable future. Aid has crowded out local revenue in the past. Once we factor out the benefits to public treasuries of recent high commodity prices, government revenue has been stagnant in sub-Saharan Africa for more than 25 years. (See http://www.imf.org/external/pu...../gupta.htm). Can Western aid donors please either help offset the unintended damage they have done, or step out of the way and let Africans – assisted perhaps by their new Chinese technical advisers – do the job.