Time for Radical Thinking? Cash on Delivery Aid (COD Aid): Taking Results-Based Financing (RBF) to the Limit

By: Lindsay Morgan¹  {photo by Shehzad Noorani}

“Ten years is a long time in anyone’s life. As time is the most valuable thing we have, because it is the most irrevocable, the thought of any lost time troubles us when we look back.”²

It’s 9:00 a.m. in Dar es Salaam and I’m sipping coffee when a friend, a nutrition specialist, taps my shoulder and hands me a report describing the state of nutrition in Tanzania.³  Over the past decade, over 600,000 children under the age of five are estimated to have died as a result of malnutrition in the country, in part because their diets lack basic nutrients to build strong immune systems. Though food fortification and exclusive breastfeeding are affordable ways to reduce malnutrition, the Tanzanian agencies responsible for their promotion have spent most of the last decade in meetings and trainings, equivocating and delaying. If Tanzania would introduce fortified flour and cooking oil, the authors of the report estimate there would be 6,700 less child deaths per year. But for now, 43,000 children will continue to die in the country each year because they are malnourished.

Later that afternoon I meet a new colleague in the lunch room.

—What work do you do, I ask?

—Private sector development, she says.

—Oh, how interesting. What’s happening with that?

—Not much.

And over dinner, a friend tells me that an agriculture program she is helping to manage, to help poor Tanzanian farmers purchase inputs like seeds, is in a shambles. “It’s just not working. But obviously that doesn’t mean we’re going to stop it.”

***

Time has not been lost. Over the last decade, aid to poor countries increased massively—by 2007, it amounted to more than $120 billion a year—and it has done huge amounts of good.

But not nearly enough. It has been ten years since the international community committed to help developing countries achieve the Millennium Development Goals, but despite the political and financial big push, many countries are still falling tragically behind.

And there is little accountability for failure—indeed, little ability to even demonstrate failure (or success) because most programs are not rigorously evaluated and data is shoddy.

Much good has been accomplished—but looking back, more could have (and should have) been done. And this fact, combined with an eroding constituency for aid nudged on by the global recession, demand evidence not only that aid is not wasteful, but that it works.

What is COD Aid?

Cash on Delivery (COD) Aid is a new(ish) idea to improve aid that is at its core startlingly simple: donors pay a fixed, incremental sum for verified outcomes and recipient governments decide how to achieve the outcomes and how to spend the reward.

Developed by experts at the Washington, D.C.-based Center for Global Development, the concept is getting noticed. Former Secretary-General of the United Nations, Kofi Annan says that: “the Cash on Delivery Aid approach could breathe new life into donor commitments to the education Millennium Development Goals and could serve as a vehicle for serious additional funding.” The Financial Times’ Undercover Economist recommends reading the book about it, and foreignpolicy.com says that it “offers accountability for donors, autonomy for recipients, and transparency for citizens of both countries.”

The donor world is also buzzing. The governments of Malawi and Liberia have issued formal written requests asking for assistance to convene donors to see how COD might be applied to primary education, and Tanzania and Ethiopia have expressed high-level interest as well. The UK government is firmly committed to COD, and the African Development Bank is considering applying it to water, while others are considering how it might be applied to HIV prevention and maternal health.

The core of COD Aid is a contract between funders and recipients. Donors agree to make incremental payments for achieved outcomes that are both measurable and continuous (thus making it possible to reward incremental progress), and verified by an independent agency.

According to the COD manifesto, Cash on Delivery: A New Approach to Foreign Aid: “At no point does the funder specify or monitor inputs. There are no required policies, training programs, or outside consultancies; no agreed contracts for building, renovating, or maintaining bricks and mortar; no specified forms of management, reforms, or decisions.” Recipient countries are free to spend the reward as they choose.

In Cash on Delivery, Birdsall and her co-authors discuss how COD Aid builds on other results-based aid programs, while going further than any other approach.⁴  COD is unique, say the authors, for combining five features that together pack a punch:

• Payment for outcomes, not just inputs OR outputs⁵
• Hands-off, recipients have complete discretion over both how to achieve the goal and how to spend the reward
• Independent verification of results
• Transparency through public dissemination of the contract and progress made
• COD aid can complement other aid programs

Is COD Aid Results-Based Financing (And Does It Matter?)

It’s hard to sell a new idea with an old name. This may be why the community pushing the move from input- to output and outcome-based aid has inspired an alphabet soup of initiatives: RBF, PBF, P4P, PBC, OBA, CCTs, and now COD.⁶ 

There is considerable confusion about both how these terms differ and why the idea behind them—of being performance-based—is really new.

In terms of the latter, should not programs that provide vouchers to women who deliver in health facilities; GAVI’s incremental payments for immunized children; and the European Commission’s variable tranches “qualify” as performance-based? What about SWAps and budget support? What about NGOs that risk losing future grants and contracts if they do not meet donor expectations? Whenever a funder pays for progress, performance, results, does this not qualify them to be included in the RBF/PBF/whatever community?

Distinctions between terms partly reflect a desire to brand ideas, but they also reflect real differences in method, and in perceptions, between them. For example, the Global Partnership on Output-based Aid contracts with service providers, as opposed to governments. PBF tends to be associated with programs in French-speaking countries in sub-Saharan Africa, and RBF, with supply-side incentive programs to health facilities and providers, though neither is in principle confined to either. CCTs are demand-side incentive programs that offer cash to families conditional on certain actions taken; they focus interventions at the individual level. Other programs focus incentives to health facilities and providers; performance-based contracting targets NGOs; and COD Aid, central governments.

Some proponents are keen for their particular slice of RBF to remain distinct, but the many terms and acronyms ultimately describe the same thing: attempts to move away from input-based financing to paying for outputs and outcomes. 

“The evangelists are looking for small theological differences when they are all trying to look at the same means to salvation,” says Philip Musgrove, deputy editor for Health Affairs. “All are looking for a way to make aid more productive.”

William Savedoff, senior fellow at CGD and Birdsall’s co-author, agrees: “Ten years from now people may not be talking about COD Aid,” he says, “but the idea of linking incentives to behavior change is definitely here to stay.”

Risky Business

Perhaps what separates COD Aid most from other RBF approaches is the level of risk it is willing to assume in order to fundamentally realign incentives in the aid system. Sit down with Nancy Birdsall and her colleagues and you might mistake them for river boat gamblers: they are just dying for you to call their bluff. Except it is not a bluff. COD Aid is not simply a thought piece to generate discussion: it is a gutsy, impassioned attempt to shake the foundations of aiddom.

Focusing on a single outcome is the first thing that strikes many as unfathomable. Says Peter Harrold, World Bank Director for Central Europe and the Balkans: “It is absolutely remarkable to have one indicator that attempts to encapsulate results in a sector.” Compare COD’s one (maybe two) outcomes to the national RBF scheme in Burundi, which has 154 composite quality indicators alone. Ditto for Rwanda, whose quality checklist contains 118 composite indicators with 1,058 data elements. Though the two schemes are different—with COD focusing on donor-recipient negotiated incentives from donor to government versus Burundi and Rwanda’s focus on donor-recipient negotiated incentives to health facilities and providers—it is unique for any development agency to focus on a single outcome.

Savedoff believes that measuring performance against many indicators, as opposed to just one, makes sense for managing health services within the health sector—he cites a New Yorker article by physician journalist Atul Gawande that shows that simple checklists in intensive care units can markedly reduce deaths. But that kind of detail should be left to healthcare managers, not donors or central governments, he says. Musgrove agrees: “If you try to measure too much you get arbitrary numbers. Less is more.”

COD also tests the limits of donor commitment to country ownership and hands-off aid.  “Where the funder is hands-off,” the book reads, “the recipient has complete discretion and responsibility. This extends from the initial design and planning right through to the implementation of strategies. All decisions and plans, including whether to have a plan, are up to the recipient government. Further, the funds a recipient receives after making progress can be used in any way, determined by the recipient:”

"The fundamental problem,” says Birdsall “is how to maximize the chances that your money will be used well without looking over the recipient’s shoulder, pretending the recipient country is in the drivers seat when in fact you are grabbing the wheel. [But] the system is more interested in completing the transfer of money than in knowing if the money made any difference. Donors are really terrified of nothing happening.”

Savedoff agrees: “If something doesn’t work now, no one is embarrassed and the data is generally not good enough that you can really say anyway.”

But the concerns about COD do not revolve solely around if countries fail. What if countries succeed—using means that donors find difficult to support?  At a World Bank brownbag lunch on COD Aid in June 2010, panelists and audience members asked whether donors can afford to “not care” about the means for achieving results. “It may be hard for the reputations of the institutions not to care,” says Harrold—about what kinds of procurement methods are used, or whether the means cause environmental damage, for example.

Emmanuel Jimenez, World Bank Sector Director for East Asia and the Pacific, suggests that COD might work beautifully in middle-income countries with robust information management and other systems, but asks: how do you address capacity constraints in low-income countries? At the lunch, one discussant suggested that the independent verification could also verify the process by which the results were achieved.

There is pressure, in other words, to insulate the recipient from both bad behavior and failure. But, as Musgrove notes: “an incentive that no one gets punished for isn’t a very good one.”

But will the promise of more money inspire or enable recipient governments to do more with existing aid modalities and policies and plans? Since COD offers no money up front, how will countries achieve new and better outcomes by old means?

And rather than inspire innovation, will COD aid simply encourage governments to do whatever is easiest to achieve the goal? Owen Barder, a Visiting Fellow at the Center for Global Development and Director of aidinfo.org, agrees that it is important to ensure that the poor and marginalized have access to services and do not have their exclusion reinforced, but counters that, in the case of education, “we should be aiming to help every child to go to school and get a quality education. If there were “cherry picking” under cash on delivery aid this would be a good thing because it would increase the number of people who benefit from services and it would provide surplus revenues which could help to pay for services to poor and marginalized communities.”⁷

But Davidson Gwatkin, a senior fellow at Results for Development, has his doubts. He says this logic makes sense if the easiest are reached only in the first phase of a program, but many programs do not continue past the first phase or donors lose interest. The hardest to reach, that is, the poorest, must be considered up front.

Of course, the reverse is also possible: COD Aid could create an incentive to work in the poorest countries or areas, where returns might be the greatest.

The other feature that takes COD to the limit is its ambition, not only to achieve clear results in a given sector, but also to inspire broad societal transformation and reform. All RBF schemes try to do this to a certain extent. For example, supply-side schemes in health, as well as trying to improve access and quality of health services, also aim to improve the entire health system by, for instance, strengthening information management systems.

COD Aid hopes for even broader reform: by requiring governments to make the terms of the contract, as well as their progress, public, COD aims to provide fodder for citizens to engage and hold their governments accountable.

The potentials of COD Aid are many, but as the British aid agency CAFOD notes, “It is critical that COD scrutinises its potential unintended consequences, and assesses the quality of outcome measurements –not simply whether or not targets have been hit.”

Vision Over Visibility

COD Aid is different from other RBF initiatives in at least one way: it has never been tried. But despite uncertainty about how it will work in practice, many believe it should be experimented with. Peter Harrold goes further, saying:“I believe we should embrace this as a fundamental concept.”

Is it risky? You bet. But the architects of COD believe that now is the time for smart risks. “Why do we think that donors know what to do?” asks Birdsall. “Do countries know more than donors? No, they probably don’t, but they are the ones who have an incentive, and responsibility, to figure it out.”

[1] Lindsay Morgan is a policy analyst and freelance writer based in Dar es Salaam.
[2] Letters and Papers from Prison, Dietrich Bonhoeffer, New York: 1953, p.3.
[3] Malnutrition: Can Tanzania afford to ignore 43,000 dead children and Tshs 700 billion in lost income every year? Uwazi, http://uwazi.org/index.php?i=278. 
[4] Take the example of GAVI, which pays countries a fixed amount for each additional child immunized. GAVI differs from COD Aid in that it provides a basic payment up front and links only  part of the funds to the outcome. In addition, the payment to recipient governments by GAVI is intended to help cover the costs of the immunization program, not to be used at the government’s discretion in other sectors or programs. Another difference is that the GAVI arrangement relies on the country’s reporting system to establish progress once that system has been vetted for reliability and accuracy, as opposed to COD Aid’s independent audit. Although the country reporting systems are vetted ahead of time, problems have been discovered after the fact.
[5] For more on COD’s focus on outcomes, see Chapter two, A Solution: Cash on Delivery Aid, in Cash on Delivery: A New Approach to Foreign Aid (Washington, DC: Center for Global Development, 2010). In addition, according to a forthcoming paper by Philip Musgrove, “Production means using inputs to deliver outputs.  These are variously known as services or interventions and include examinations, tests, surgeries, consultations and treatments of all kinds, including hospital stays paid by the day.  Outputs are intended to yield outcomes that benefit the patient.  Good prenatal care should mean babies born with an adequate weight….”
[6] Results-based financing, performance-based financing, pay for performance, performance-based contracting, output-based aid, and conditional cash transfers. 
[7] See Cash on Delivery Aid: Response to CAFOD Questions, http://www.owen.org/blog/2864. Barder also co-authored the CGD working paper "Payments for Progress: A Hands-Off Approach to Foreign Aid," which originally outlined the COD aid concept.