DevExplains: Are we overcounting ODA by tens of billions?

Official development assistance, or ODA, is the primary measure of development spending by the world’s largest donors. Preliminary figures say that $178.9 billion was spent last year. But what does that number really mean? How accurately does its measure of “donor effort” reflect the real contribution of high-income countries to development in low- and in middle-income countries? Not very, is the answer from two experts who have studied ODA closely. They say that over the years, the Development Assistance Committee, the arm of the Organisation for Economic Cooperation and Development that makes the rules about what counts as ODA, has been bullied by donors into accepting more and more things as ODA that were not in the spirit of the original rules. One of the critics is Euan Ritchie, a senior development finance policy adviser at Development Initiatives, an organization that gathers and analyzes data on the development sector. When ODA was first adopted as a measure, it was intended to track capital flows from high-income countries to lower-income countries, he tells Devex. “But since then, the DAC members who are responsible for most of these flows have tried to gain credit for more and more things,” he says. “Foreign secretaries want to go to these big international conferences, and brag. And so there was an incentive to count more and more things.” He identifies two big issues: loans and in-donor refugee costs. Loans currently amount to around $12 billion of ODA per year, but if they were counted more fairly, he says, that would be reduced by around 80% — knocking about $10 billion off the value of ODA in a typical year. Refugee costs, meanwhile, average another $12 billion a year, albeit with wide year-on-year variation. They are purely domestic spending, and do not contribute at all to development in low- and middle-income countries. There are growing calls for DAC to change how ODA is measured, to more accurately reflect how much high-income countries are really spending. And there are also calls for DAC itself to reform, to make sure it is more independent. Perhaps the leading voice for change on the subject is Steve Cutts, a former assistant secretary-general at the United Nations and former chef de cabinet at OECD, who described the current data as a “corrupt” set of statistics. Cutts is highly critical of DAC’s lack of independence from its 31 donor members. “The dogs are in charge of the dog show, and they are all giving themselves rosettes at every turn,” he says. So what are the problems with the figures, and what is to be done about it? The problem with loans Since 2018, OECD has calculated ODA on a “grant equivalent basis.” When donor countries make loans to low- or middle-income countries, they can count only the grant equivalent element of any loan as ODA. To calculate grant equivalence, the DAC uses a base “discount rate” of 5%, and then adds between 1% and 4% to that rate, depending on the World Bank income category of the country receiving the money, to cover the risk of nonrepayment. But Cutts says this is not a fair rate for concessional interest rates. He says since ODA is meant to measure “donor effort” — the amount of resource actually given up by the donor — a loan should only count as aid if a country lends at less than its own cost of borrowing. Typically, for DAC members, the cost of borrowing is far lower than 5%. Which means that right now, DAC members can borrow money on the bond markets, lend it at a profit, and have it count as aid. The need to include a risk premium would be more reasonable, he says, if donor countries faced a genuine risk of taking a loss on their loans. But when a low- or middle-income country does default on a loan, the donor is allowed to count the cost of any rescheduling or write-off as additional ODA — a situation condemned by Cutts as “double counting.” Ritchie says OECD already uses a different system to measure the concessionality of loans in another part of the organization — the Differentiated Discount Rates, which are based on the cost of government borrowing. This system is used to measure the concessionality of trade-related tied aid as part of disciplines to restrict the use of aid that requires the use of goods and services from the donor country. If DDR was used to calculate the grant equivalent value of loans it would reduce the amount counted by 80%, he says. “The OECD — a different part of the OECD — has been measuring the grant equivalent of loans quite well, for a long time,” Ritchie says. “When they decided to measure the grant equivalent of loans, I think lots of people just assumed that they would use that same method. But that didn't prove palatable for the DAC members.” Refugee costs Donors are also permitted under OECD rules to count the cost of supporting refugees who arrive in their countries as ODA. These rules only permit certain costs to be counted, and only for the first year, but nonetheless, the costs can be considerable, as has been highlighted by soaring refugee costs from the war in Ukraine. Right now, the United Kingdom government is facing internal criticism for counting the whole cost of housing and feeding Ukrainian refugees as ODA, with the counterintuitive result, according to BBC reports, that more bilateral aid is being spent within the U.K. than outside it. A joint tracker from the ONE Campaign and Donor Tracker estimates that refugee costs will exceed $49.6 billion this year — an estimate that has been rising steadily for months. In previous years, refugee costs as a proportion of ODA have varied from $6.6 billion in 2014 to $18.5 billion in 2016 in constant 2021 prices, averaging just under $12 billion. The arguments over refugees are more nuanced than those around loans, Ritchie acknowledges, because there is a duty of all countries to care for refugees. However, his position is that on balance, these costs should be excluded from ODA because they are not spent with the economic development or welfare of low- and middle-income countries as their main objective — something which he points out is a key part of the definition of ODA. What now? Cutts tells Devex that part of the problem is that the DAC is not independent of its 31 donor members. Though there are consultation mechanisms with low-income countries and NGOs, Cutts says that such actors are not part of the decision-making process on what counts as ODA. And he says the DAC needed to be “much freer from political pressures than it is now.” Part of the problem, in his eyes, is that the chair of the DAC is up for election every year. “You give me a job for five years, then I'm independent,” he says. “You tell me I'm going to be subject to reelection every year [and] you're completely dependent on your members.” Cutts also took aim at the secretariat of the DAC, staffed by OECD officials: “I don't know what resistance it may have given in the past,” he said. “Its job now is to try and justify every committee decision that's been made … to kind of almost say to everyone ‘look, nothing to see here.’” Cutts penned scathing critiques of the DAC rules in the Financial Times and Brookings earlier this year, prompting the secretariat to publish a set of frequently asked questions on the modernization of ODA. Cutts dismisses the secretariat’s response however, as failing to respond to his critique. It was “deliciously ironic.” Cutts says that OECD was the organization that published “Twelve key recommendations to ensure the quality of statistics,” even while ignoring its own advice in the case of the DAC on the need for things like impartiality, objectivity and transparency, and sound methodology. Asked what he’d like to see now, Cutts says he wants to see an independent statistical review of ODA. Ideally, the OECD itself would conduct any review, though if that proved politically unpalatable, Cutts says other organizations, such as the World Bank, could also do the job. The challenge of reviewing ODA rules outside the auspices of OECD is “whether or not it will be given any credence or credibility,” Cutts says. “I don’t know.” But without such a review, Cutts argues that OECD’s deserved reputation for sound statistical work and policy analysis risked being undermined as it is “effectively being used to launder these corrupt statistics from the DAC.” OECD was approached for a response but declined to comment. Update, Dec. 5, 2022: This article has been updated to clarify Euan Ritchie’s position on why refugee costs should not be counted as ODA. Update, Dec. 9, 2022: This article has been updated to clarify that Differentiated Discount Rates are used to measure the concessionality of tied aid.