OECD response to critiques of its ODA changes
A set of FAQs, dated September 2022 and entitled “The Modernisation of ODA: Frequently Asked Questions”, comprise the only substantive written public response offered to my (and others’) critiques of the Development Assistance Committee’s (DAC) changes to ODA.
A careful analysis of the FAQs shows that they are totally inadequate and unconvincing. They do not discuss many of the flaws raised, and on the points they do address, the responses are evasive and largely irrelevant. I therefore sent a full rebuttal of them to the Organisation’s Deputy Secretary General. Noting that the FAQs claimed to have been drafted “with a view to promoting a transparent, informed and constructive dialogue”, I requested that my responses be posted on the OECD website – but they were neither posted nor even acknowledged.
Drawing on my rebuttal, I summarise below the main inadequacies in the OECD’s responses to the questions it asked itself. I then point to a few important technical questions that the FAQ did not ask.
The OECD's document does not start well. It explains its existence by first outlining the DAC’s “ODA modernisation” process and then stating that
This process has sparked a debate among policymakers, statisticians, civil society organisations and other observers who value official development assistance (ODA) as the ‘gold standard’ of foreign aid.
In fact, the DAC and the OECD Secretariat have NOT debated the issues but have preferred instead to ignore the growing criticisms, issue obfuscatory statements and attack their critics. And frankly the claim that ODA is the “gold standard of foreign aid” has been stripped of any credibility.
FAQ 1. Why change the way of counting ODA?
The OECD’s answer here is that moving to grant equivalents provides a better comparative measure of the concessionality in loans. That is true in principle, but it’s only true in practice if the grant equivalent calculation uses realistic discount rates. The DAC does not – and this is the whole point. Its discount rates massively undervalue the returns it gets on loans, so that even loans at full market rates appear to have grant equivalents, and other loans that are only slightly concessional are calculated to have huge grant equivalents.
FAQ 2. Was the ODA modernisation process transparent? Who was involved?
On this the DAC claims that its “modernisation” was “fully transparent…and inclusive involving many technical experts, policy makers, and other key stakeholders from DAC member countries, civil society organisations (CSOs), and beyond”. This is untrue. The origin of the recommendations finally agreed at the 2014 DAC HLM was a "High Level working group" of eight senior DAC representatives under the chairmanship of the UK DFID’s Permanent Secretary at the time, Mark Lowcock. This group met and negotiated in secret. No records of its discussions have ever been published, though faint traces of its existence can be found on the web. The FAQ’s claim that, during this process “all papers were circulated as ‘unclassified’ documents” is also untrue. The Lowcock group circulated no papers at all, and two of the four earlier DAC papers to which the FAQ itself provides links (FAQ 6) are still stamped “Declassified”, showing that they were originally circulated as classified.
FAQ 3. Has the measurement of ODA deteriorated? Has the quality of aid declined?
These are two very different questions. On the first question, the rest of this website shows how the measurement of ODA is flawed in multiple ways. The rules introduced in recent years score items that involve no real fiscal effort, over-score the fiscal effort in others, introduce alternative ways of scoring the same items, arbitrarily cap the reporting of return flows, etc. The rules are also over-complicated and contain internal contradictions. The FAQ dodges all these points and instead asserts that the new system is better because it corrects inconsistencies in interpretations of which loans score as ODA. But the problem is that the new “quantitative definition of concessionality” sets the bar far too low, so that consistency is achieved only by ODA scoring of many unsubsidised, non-concessional loans. On the second question, the new rules introduce perverse incentives by scoring loans higher than grants for the same fiscal effort, and never scoring any net negative amount for equity investment, even if huge profits are made. Over the longer term, the quality of aid is bound to decline as the DAC moves to over-score more and more non-concessional or marginally concessional items that allow donors to meet aid targets with minimal real outlay. On this issue, the FAQ just waffles, e.g. “the DAC encourages all donors and partners to adhere to the principles of country-led development, mutual accountability, inclusive partnerships, transparency, and a focus on results” – pretending that such “encouragement” is sufficient to counteract the force of the perverse incentives the DAC has itself deliberately created.
FAQ 4. Is official support for the private sector unduly recorded as ODA?
As explained elsewhere on this website, the correct answer to this question is clearly yes. The DAC Secretariat dodges it by referencing the DAC’s drawn-out negotiations on permanent rules for scoring this support and excusing the clear over-counting in its 2018 “provisional” rules by saying that the instruments involved “only represent 2-3% of total ODA”. But the point of the rule changes was precisely to provide incentives to increase the use of these instruments, and in any case, if scoring them as ODA is wrong, as it is, then changing the rules in order to score them is also wrong, regardless of the process by which this is done or the size of the resulting error.
FAQ 5. Does the grant equivalent system over-count loans in ODA? What is the impact of ODA modernisation on ODA figures overall?
Putting these two largely unrelated questions together is an attempt to make the reader think that if the “modernised” ODA grant equivalent figures aren’t much higher than its old “capital flow” figures, then loans are not being over counted. This is fallacious, because as loans currently comprise a small share of ODA (around 10-15%), even over-counting them by a lot will not make a big difference to total ODA in percentage terms. And in any case, under “modernisation”, the same low hurdle applies to both “cash-flow” and “grant-equivalent” measurement. Since both measures now count non-concessional loans, comparing the two figures is meaningless as a test of overcounting in either. In the long run, any grant equivalent system scoring only positive grant equivalents will record more ODA than a capital-flow system which deducts all loan principal payments until there is a zero balance overall. And the FAQ does at least have the decency to admit that “the introduction of the grant equivalent system will increase ODA figures in the long term”.
FAQ 6. Do the discount rates agreed by the DAC unduly inflate ODA? Would differentiated discount rates (DDRs) be a better option?
As the rest of this website and several other studies have shown, the DAC’s discount rates are way above major lenders’ cost of funds, and have thus massively inflated the grant equivalents of loans reported as ODA since 2018. The DAC’s more than 500-word answer on this question never actually denies this. Instead it tries a variety of unsatisfactory defences, including: • The DDRs “are not recognised as a universal benchmark” – Well, maybe not, but they are accepted by the same OECD countries as an accurate measure of concessionality of tied aid, and the DAC is purporting to measure the same thing. • “The 2014 agreement on concessionality was the result of transparent, in-depth discussions among technical experts, policy makers, CSOs and more stakeholders, who carefully examined all options” – Untrue, see FAQ 2 above. • Use of “volatile” rates would lead to “lack of predictability in ODA budgets”. This is not true either: in fact, the opposite is the case. If the DAC used current market rates, they would score predictable ODA credit for budgeted subsidies - ideally, they would score the exact amount of the subsidy. By contrast, the DAC's own fixed discount rates mean that any budgeted subsidy applied to a loan made later from funds raised later on the market will produce an unpredictable amount of ODA. • “The impact of the choice of discount rates on the valuation of loans is relatively low.” This is patently false. The choice of discount rates is precisely what determines the value of loans in a grant equivalent system. You can get any grant equivalent you like by just varying the discount rate.
FAQ 7. Is risk double-counted in the reporting on ODA loans and ODA relief?
The correct answer is yes. By including a risk margin in the discount rate used to calculate the “grant equivalent” of the original loan, a donor claims ODA upfront for effectively insuring each loan against default. It assumes in its reporting of EVERY original loan that some will not be repaid in full, and it “banks” ODA in advance for the additional concessionality it may have to offer by way of relief on those that go bad. So claiming more ODA on those that do go bad and require debt relief is clearly double counting. The FAQ attempts to obscure this point, claiming that “There is no double-counting: only additional concessionality involved in debt restructuring is counted as ODA.” The basis of this claim is that the DAC nets out the original ODA claimed upfront for assuming risk when it scores ODA for debt restructuring. But it only deducts this “original ODA” for restructured loans; it leaves in place all the ODA it has scored on other loans where the risk did not materialise. To illustrate the point, one can use a simple analogy of a property insurer that has a portfolio of 100 houses dotted around the country. Let’s say that by the end of the year, the insurer expects to have claims amounting to no more than 2% of the total value of the property insured. On that basis, the insurance company can reasonably charge a 2% premium on each policy to cover the overall risk and ensure its own commercial viability. What the insurance company cannot then do is to visit a householder the morning after their house burns down and claim that overnight the risk increased because their house caught fire, and that therefore the householder has to pay an additional insurance premium to cover the other 98% of the value of the property. Yet this is what the DAC is doing in claiming ODA for the “additional concessionality” (i.e. the value of relief) associated with a loan default. The donor has already booked ODA credit for effectively self-insuring its whole portfolio of loans, so it cannot claim additional ODA just for those loans that are not repaid in full. That is clearly “double-counting” the risk. Looked at another way, donors could only justify scoring extra ODA for the “additional concessionality” in relief if they were also to score negative ODA for all the loans that were repaid without problems and where the extra ODA scored upfront for risk therefore turned out to be unwarranted….but in that case it would be a far neater solution to just score ODA for relief when it occurred and remove the credit margin in the upfront discount rate used to calculate the ODA grant equivalent of every loan. (Incidentally, the ceilings on counting such additional ODA credit for relief that are cited in the FAQ as “strong safeguards” would, in the analogy of the property insurer, only limit and not eliminate the double-counting, by capping the amount demanded after the fire to the full original value of the house minus premia previously paid.) When the DAC decided to introduce grant equivalent reporting on loans in 2014 it promised that “Given that the new system would value upfront the risk of default on ODA loans, the eventual forgiveness of these loans would no longer be reportable as a new aid effort.” The FAQ fails to mention this promise or give any valid reason for breaking it.
FAQ 8. Does the new methodology encourage donors to provide loans rather than grants, potentially worsening the debt crisis?
Given the very low interest rates in major currencies in 2018-21 – far lower than the DAC discount rates – the system has been scoring non-concessional loans as ODA and massively inflating ODA for loans that are only marginally or modestly concessional, and so has given donors a big new incentive to provide loans rather than grants, thereby increasing recipients’ debt burdens. The FAQ’s defence is “At present, analysis reveals no sign of major changes in donor behaviour: grants continue to represent around 80% of total ODA.” But donors take time to alter their programmes, and even if the loan share was not rising, it would not prove there was no incentive to do so, only that the incentive had not yet had much effect. However, 2022 figures show a big increase in loans which had already risen since the new rules came into effect, largely negating the FAQ’s argument.
FAQ 9. Does the reform affect the composition of climate finance negatively? Are OECD figures inflated?
The FAQ starts out by claiming that “Climate finance figures are largely unrelated to the grant equivalent system adopted by the DAC” but this has not been true since the UNFCCC has started to collect grant equivalent figures on climate finance. The FAQ also tries to pretend that climate finance is mainly grants by focusing only on “bilateral concessional climate finance” which is only a small share of the total. It repeats the line that “The introduction of the grant equivalent system in DAC statistics has not induced an increase in the share of loans in ODA”, which as noted above is belied by the 2022 data. But bilateral donors are in a unique position to provide the grants that all parties agree are desperately needed, as too much climate finance is currently being extended in the form of loans. Yet the proportion of grants is significantly lower for climate finance than for other bilateral aid. And the DAC’s exaggeration of donor effort in loans over grants provides a powerful incentive against changing this. Moreover, the terms of the ODA loans being extended are barely concessional (when the grant equivalents are calculated using robust discount rates).
FAQ 10. Does the methodology over-score ODA on relief for non-ODA debt?
Again, the correct answer to this question is Yes. The OECD FAQ again evades the question, this time by noting that the practice of counting ODA for debt relief for non-ODA loans was not introduced by the 2014 “modernisation”. No one has claimed otherwise. However, this counting had been criticised on numerous occasions in the old “capital flow” system for ODA, and it is indefensible in a “grant equivalent” system that lacks reverse-flow entries for the cancellation of loans, and that explicitly claims to represent real donor effort. It was therefore incumbent on the DAC, when it introduced the grant equivalent system, to examine whether there was any real effort in relief on non-ODA debt. Instead, the DAC Secretariat simply asserted in its new instructions that since no grant equivalents would previously have been reported for a non-ODA loan, its forgiveness should be recorded as a new grant equivalent, and that the amount of this grant equivalent should be the total amount forgiven. The instructions thus completely ignored the question of whether there was any real donor effort. The FAQ now tries a different justification. It points out that debt relief for export credits provides a benefit for developing countries. But this is beside the point. The DAC has defined grant equivalent ODA as a measure of donor effort, not recipient benefit. Only real donor effort could justify scoring grant equivalent ODA for relief of non-ODA loans. But there is no donor effort involved in debt relief for export credit debt (or similar loans) where risk-related premium fees or loan margins have been levied from developing country borrowers (or commercial exporters) to cover those same costs, as is required under WTO rules.
FAQ 11. Buying shares in financial structures counts as aid but selling them is not deducted. Does the treatment of private sector instruments unduly increase aid figures?
Here, even the question contains an error. Equity sale proceeds are deducted; it is just that the amount deducted is capped at the original purchase price, so that ODA figures hide any profit that has been made and thus overstate net financial effort. This does indeed increase aid figures unduly, since loss-making shares in financial structures score ODA, while profit-making shares can never score negative ODA, no matter how much money the “donor” makes.
FAQ 12. Are rich countries sending doses of vaccines purchased for their own populations – in some cases outdated – to poor countries, and counting them as ODA?
The FAQ’s answer effectively concedes that unused second-hand vaccines were passed on to developing countries and scored as ODA, though it points out that “Expired doses are not eligible”. However, the controversial question here has rather been whether donors have scored more ODA for donating vaccines than they actually paid. On this, there is still a lack of full transparency. See our detailed paper on this.
So, as a rule, the FAQ dodges rather than responds to the questions it asks itself, and in several cases it asks the wrong question. But perhaps more importantly, despite claiming that it “addresses the technical aspects underpinning this debate”, it avoids even asking many other embarrassing technical questions, including the following, all of which have been raised by myself and/or others in public over the last couple of years:
Can donors now score more ODA for forgiving an ODA loan than the amount they forgive? (Correct answer: Yes)
Can a rescheduled loan eventually score as much as a grant, even if it is fully repaid? (Yes)
Can a donor score more for forgiving a non-ODA loan than the loan was worth in the first place (Yes)
Can donors pick and choose whether to report as ODA what they give their development finance institutions, or what the institutions lend or invest in developing countries, and even switch between the two methods? (Yes)
Did the DAC admit in 2016 that “private sector instruments” were “non-concessional in nature” but then spend nearly three years trying and failing to work out how to count them as ODA anyway, and is it now restarting this process? (Yes, Yes and Yes).