Print
Share
What Is the Counterfactual for COD Aid?
July 15, 2010
This is a joint post with Nancy Birdsall.
We often hear criticism of the COD Aid approach from people who question whether a high-level incentive would really alter the behavior of recipient countries. Paolo de Renzio raised this issue in a recent blog, saying that COD Aid is unlikely to work because recipient “[g]overnments have not only insufficient capacity, but also limited political interest in using available resources to maximize development impact.” The question this raises, though, is not whether COD Aid is worth trying. Rather it is questioning whether any foreign assistance can make a difference. The appropriate counterfactuals for COD Aid are existing aid modalities which rely on extensive engagement between funders and recipients on inputs, planning and implementation. The assumption of current aid modalities is that imposing the funder’s views of planning, institutional structures, training, and technical assistance from abroad can achieve progress even when a recipient country is less than enthusiastic about a program. That is the counterfactual against which to consider COD Aid.
The key thing missing from these critiques is that they implicitly treat aid to low-income countries as flowing from a beneficent donor to a problematic recipient. However, as Paolo and others are well aware, the behavior of funders also contributes to failures in achieving development goals. Funders don’t necessarily know better than countries how to achieve results, yet the structure of technical assistance tied to aid projects assumes this to be the case. Also, funders introduce volatility in funding, information requirements, and policy attention that often detracts from domestic governance in these countries.
Nancy and I are not arguing that COD Aid is worth trying because it creates a better incentive for recipient countries. We are arguing that it is worth trying because it creates a better relationship between funders and recipients. It would focus attention on the desired outcome, on getting precise and reliable information about that outcome, and on directing funds in proportion to progress. It would shift variability in payments away from political and bureaucratic processes in the funding institutions and toward factors related to achieving progress that are more in the purview of the developing country. It changes the structure of information reporting and payment triggers for both funders and recipients in a way that has not been fully tested anywhere.
In the end, if COD Aid to a country with no interest in achieving progress is like pushing on a string, at least it brings no further burden on the developing country and doesn’t waste developed country money. By contrast, most current aid modalities disburse money to a government that isn’t making progress, creating vested interests in maintaining the flow of funds even when the outcomes are not achieved.
Share 
You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
7 Responses to “What Is the Counterfactual for COD Aid?”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
July 15th, 2010 at 11:08 pm
While I strongly support experimentation in COD schemes for all the reasons you give, I don’t think one can ignore the issue of “problem recipient” countries, even if no donor country is problem free. Surely we should prioritize countries for COD implementation on the basis of demonstrated will and ability to use cash effectively. You seem to be saying that if COD payments are used ineffectively, at least we will have avoided all the bureaucratic hassle that is inherent in the existing delivery mechanism for ineffective aid. This may be true, but it is hardly a compelling argument. Donor countries have long demonstrated their willingness to pay more to maintain more control over the funds, and have a better sense of where their money went even when the desired outcomes were not achieved.
July 16th, 2010 at 9:29 am
I appreciate your support for trying COD schemes, but think that some of the concept may not be entirely clear based on your concerns.
First, COD Aid does not have to be prioritized for willing countries because the money only goes to countries that make progress. Rather than determining beforehand which countries are “deserving” of support, the funder offers to pay in proportion to progress on the openly agreed goal. The quality of the government is revealed by their progress rather than by their promises.
Secondly, I probably wasn’t clear enough if it sounded like I was saying “that if COD payments are used ineffectively, at least we will have avoided all the bureaucratic hassle…” Rather, I’m saying that a COD contract avoids the bureaucratic hassle both (1) when a country successfully makes progress and gets payments AND (2) when a country fails to make progress and does not receive payments. In this sense, COD payments by design cannot be used ‘ineffectively’ since they are only paid against real outcomes.
July 16th, 2010 at 1:02 pm
Interesting post, Bill. How is this different from outcome-based aid that the World Bank and others have been piloting and promoting for a decade or so? Is it an operational or relational difference with governments, since the basic premises of COD and OBA are the same it seems to me?
July 16th, 2010 at 4:01 pm
Your additional explanations are helpful, but don’t completely allay my concerns. The process of negotiating an outcome-based COD contract and agreeing on a monitoring mechanism requires recipient countries to have some level of bureaucratic efficiency and transparency in evaluations. Would you want to bet the credibility of the COD concept on introducing a COD contract in a country that you suspect would cheat on reporting of outcomes?
In theory, the COD concept does prevent of any ineffective use of contribution. The issue will arise if an when a government with a COD contract improves outcomes through no effort of its own and simply uses the cash for purposes not related to the improved outcomes. This would be perceived as using COD funds ineffectively even though consistent with the terms of a contract. Likewise, a government that makes considerable effort to improve targets, but for reason beyond its control fails to meet the targets will feel cheated, even though this is also consistent with COD principles. Many people accustomed to rewarding efforts rather than results will sympathize.
July 16th, 2010 at 5:23 pm
This post provides a clear and concise explanation of a key aspect of COD – that it can be an improvement over other aid modalities (and should be compared to them), not that it’s a silver bullet for development. While the transaction costs of setting up a COD contract should not be underestimated (especially the first pilots), the transaction costs of current modalities are also quite high and too easily disregarded.
Thanks for another very interesting post.
July 19th, 2010 at 9:52 am
David W asks how COD aid differs from “Output-Based Aid” (OBA) at the World Bank. The main difference is that OBA is focused on outputs at the service provider level, e.g. a water company producing water services, a firm constructing roads. COD aid is formally the same – paying for a result – but is aimed at untangling the aid funder-aid recipient relationship.
Jeffrey Barnes raises two really important points.
First, he points out the difficulty of signing an agreement if there is a risk of reporting on outcomes. This is why one of the five basic features of a COD Aid agreement is independent verification of the outcome measure. In our book, we show how this would be done in an education agreement – the govt reports on the number of children who reach the final year of primary school and take a national assessment exam. An independently contracted team conducts a re-test of a random sample of schools two to four weeks later and estimates the national total and assesses the accuracy of the school-level reports. Assuming the country’s report is within a reasonable confidence interval, they get paid according to their report. If the verification shows the report was misleading, then there are penalties.
This is the same as any procurement process – government agencies and companies buy, for example, laptop computers and then they have a process of opening the boxes and verifying that they got what they ordered in numbers and quality! If there is any problem with the delivery, payment is deferred. Sometimes suppliers even have to pay for delivering shoddy products; and they can be fined or sued for misrepresenting the delivery if it is intentional.
Jeffrey’s second point is that COD Aid payments would necessarily be overpaying in cases where recipients are lucky and underpaying in cases where they are not lucky. While that is true, the comparable statement about current aid modalities is that they pay the recipient too much when it might have cost less to get the same results and too much when funds were wasted (or stolen) without achieving results.
By paying for verified results, COD Aid at least assures that payments only occur when something positive has happened. It doesn’t make sense to create a COD Aid agreement for a goal that depends more on variations in factors outside the government’s control than factors that they do control. So, this is more a practical question – probably more comfortable to affirm with regard to education than, say, with private sector employment growth.
It is a shame that recipients who try really hard without success would not get their payments. But this is the crux of country ownership and responsibility. We can only protect recipient countries from this risk by aid approaches that reduce their autonomy and introduce serious distortions in accountability and incentives.
I think that Jeffrey’s last sentence really gets to the root of discomfort with COD Aid. People who design, fund, and implement aid programs are more comfortable with rewarding effort than rewarding results, despite all the rhetoric to the contrary. Yet our ability to measure ‘effort’ is probably much worse than our ability to measure outcomes. I can construct a more precise measure from household surveys of educational attainment or health than I can of a government’s seriousness with respect to passing legislation, designing policies, staffing, implementing, etc. Having looked at audit reports, I’m much more confident that I can accurately measure outcomes than measure whether funds have been disbursed properly.
Finally, I don’t expect that these comments will allay all concerns. This is an idea that has consistency with certain principles and tackles the aid conundrum in a significantly different way than other forms. Therefore it presents different kinds of risk than other forms of aid. After analyzing the advantages and disadvantages, and from my own experience with aid programs, I think these risks are probably much lower than the ones faced by current modes and therefore worth trying.
July 30th, 2010 at 4:23 pm
Thank you Bill, for a very interesting post delving into the discussion on Results-Based Financing (RBF) and how to better assert its effectiveness as compared to other (more traditional) modalities.
On the question of what is counterfactual to COD, one would probably need to take into consideration the outcome being pursued. For example, COD operations related to high level outcomes like the implementation of certain policies in specific sectors, would call for a comparison to budget support aid operations. However, COD operations linked to lower level outcomes like improved levels of service (be it in education, water supply, etc.), would be best compared to programmatic approaches that build a chain of aid agreements where each step, after the first one, depends on the successful completion of the previous one.
In response to David W’s question on the difference between Output-Based Aid (OBA) and COD – both are part of the RBF spectrum, and aim to leave the “how” an outcome or output is achieved to the creativity and capacity of the implementing agency. In the case of OBA, there is a link between the cost of achieving the output and the amount of aid disbursed and considerable effort goes into trying to make the aid as efficient as possible in terms of the output/aid ratio.
At this point any comparison is theoretical, although one would expect that lessons and experiences from OBA and other RBF mechanisms that have been piloted could inform how COD is eventually applied. For information on OBA visit http://www.gpoba.org