How much is aid worth? Explaining the DAC’s proposed changes to the measurement of ODA

On 16 December 2014, the Organisation for Economic Co-operation and Development (OECD)’s Development Assistance Committee (DAC) finally published its proposals for changes to the measurement of official development assistance (ODA). The most significant change was in the way concessional loans are valued for ODA. The current method of valuing loans has been widely criticised as overstating the benefit of ODA loans to developing countries.

Two of the main criticisms are:

1. Under the existing system, any loans with a grant element[1] of more than 25% are included in the ODA figures in their entirety. This means that loans with low concessionality (just over 25%) are valued the same as highly concessional loans.
2. The reference rate[2] used in the calculation of the grant element of ODA loans has, for years, been set at the artificially high level of 10%. The outcome of this is an inflated grant element to the extent that even loans at commercial interest rates could be counted as ODA.

The DAC has responded to each of these criticisms by proposing the following changes to the ways loans are valued in ODA statistics:

1. Only the grant element percentage of the loan, and not the full face value, will be counted as ODA. This means, for the first time, loans with a very high grant element will count for more ODA than loans with a relatively low grant element.
2. The flat 10% reference rate will be replaced by a range of rates, based on the International Monetary Fund (IMF)’s reference rate plus a ‘risk premium’ that depends on the status of the borrower. The ‘risk premium’ has been set at 1% for upper middle-income countries (UMICs), 2% for lower middle-income countries (LMICs) and 4% for low-income and least-developed countries (LICs/LDCs). Since the IMF reference rate is currently 5%, this gives an effective reference rate of 6% for loans to UMICs, 7% for LMICs and 9% for LICs/LDCs.
3. The current 25% grant element threshold will be replaced by a range of threshold values, also based on the income status of the borrower. It is proposed that, in the case of LICs/LDCs, only loans with a grant element of over 45% should be counted as ODA. For LMICs and UMICs the threshold levels will be 15% and 10%, respectively.

The changes in policy towards ODA loans are, on balance, a very positive step. We have previously highlighted the benefits of valuing only the grant element of aid loans as ODA and the DAC’s acceptance of this is a very welcome development. The move towards IMF reference rates for loans is a step in the right direction, although it can be argued that the differentiated discount rates (DDRs), published by the OECD themselves, are a more realistic basis for reference rates. Still, at least in the case of countries classed as ‘middle income’, we have a significantly lower (and thus more realistic) reference rate. However, the proposed 9% reference rate for LICs/LDCs is only slightly lower than the old 10% rate.

The addition of ‘risk premia’ to the reference rate also creates a potential danger of over-incentivising donors to lend to LDCs/LICs. This is because, under the proposed system a loan to an LDC/LIC will lead to a greater amount of ODA being credited to the donor than for the exact same loan to an MIC. For example, a $10 million, 25-year loan at an interest rate of 2% would be counted as $454,000 of ODA if loaned to an LIC, but only $313,000 of ODA if loaned to an UMIC. The 45% grant element threshold for LDCs/LICs will mitigate this potentially perverse incentive to some extent.

The linking of the ODA rules to the IMF’s debt sustainability rules and the World Bank’s borrowing policy, which was also announced in the DAC communiqué, is a positive move and will help to protect LICs against over-lending.  It could even be argued that these safeguards are necessary due to the potential incentive to over-lend to LICs/LDCs created by the 9% reference rate.

In summary, although questions remain around the arguments for using different discount rates for LDCs/LICs and MICs, the other constraints on lending to LICs/LDCs (45% threshold and adherence to IMF debt limitation policy) provide a counterweight to any perverse incentive to over-lend to the poorest countries. Therefore these changes are, overall, welcome.

The other rules on the valuation of ODA have largely remained unchanged and this is somewhat disappointing. The DAC have failed to take this opportunity to remove elements such as imputed student costs and donor-country refugee costs, which seem set to continue to qualify as ODA. As we highlighted in our 2013 Investments to End Poverty Report , these two elements alone accounted for over $6.5 billion of ODA reported by DAC donors in 2012, but resulted in no actual transfer of resources to the developing world. A future broader measure of development finance the OECD proposes may be a more suitable way to report these items rather than including them in ODA.

[1] The grant element is the standard way of measuring how concessional a loan is. It can be viewed as the difference between the cost, in today’s prices, of the future repayments a borrower will have to make on the loan in question and the repayments the borrower would have had to make on a non-concessional loan. This is therefore the amount of money that is considered to have been ‘given away’ by the donor, hence grant element. The grant element is normally shown as a percentage of the value of the loan.

[2] The reference rate is a benchmark interest rate used in the calculation of the grant element – the higher the reference, the higher the grant element. For an accurate measure of concessionality, the reference rate should have some relationship to the market rate faced by the lender.