With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
Cash on Delivery is an approach to foreign aid that focuses on results, encourages innovation, and strengthens government accountability to citizens rather than donors. Under COD Aid, donors would pay for measurable and verifiable progress on specific outcomes, such as $100 dollars for every child above baseline expectations who completes primary school and takes a test. CGD is working with technical experts and potential donors and partner countries to design COD Aid pilots and research programs.
Cash on Delivery Aid is designed to overcome the problems of traditional aid, which often focuses more on disbursements and verifying expenditures than on results, undermines a government’s accountability to its citizens, and undervalues local experimentation and learning. COD Aid’s advantage is in linking payments directly to a single specific outcome, allowing the recipient to reach the outcome however it sees fit, and assuring that progress is transparent and visible to the recipient’s own citizens. These features rebalance accountability, reduce transaction costs, and encourage innovation.
COD Aid can be applied to any sector in which donors and recipients can agree upon measurable, verifiable outcomes and commit to making progress toward those shared goals. The approach is fully explained in Cash on Delivery: A New Approach to Foreign Aid (CGD, 2010). Listen to more about COD Aid in these Wonkcasts. Explore the links to the right for more information on specific sectors and countries.
Get Cash on Delivery Updates
CGD experts offer timely research, analysis, and policy ideas for the world’s emerging development challenges. Sign up to get the latest updates from CGD!
Last week I gave a seminar at DfID in London on COD Aid. The turn-out was great—standing room only, with people from overseas offices videoconferencing in. Here is the PowerPoint I used.
Now that the leadership of DfID is committed to trying COD Aid (at least three pilots are promised), the discussion in the official aid world is switching from “Why COD Aid is Hard to Do” (‘and could you please tell me about some real-live successful examples before we try it’) to “How to Do COD Aid Well.”
When you come to the slide of the original Boston Tea Party picture, keep in mind the following: COD Aid is meant to address the problem that in aid-dependent countries (where aid replaces tax revenue) there is often ‘No representation without taxation’—a reversal of the original tea party movement’s cry. For web readers I inserted an additional slide into the web version of my presentation to explain.
Seminar participants had great feedback and good questions. Chris Berry, from DfID’s Ethiopia office, explained the current discussions his office is having with the Ethiopian government about implementing a COD Aid pilot. Rakesh Rajani, who heads Twaweza in East Africa, referred to COD Aid as an antidote to Tanzanian’ cynicism about aid; in his words, “Citizens in Tanzania are cynical too…. Aid is seen as gravy to insiders.” And Andrew Rogerson of the OECD (who has a terrific review of COD Aid here) said something like, “It is time to wrap our brain around the idea they (the recipients) will fix things in ways we (the donors) don’t.”
Both they and others had some critical points to offer as well. If you attended the meeting in one form or another please feel free to add your own questions, concerns, and complaints below or send them to me via email.
In a presentation delivered at the UK Department for International Development on March 9, 2011, CGD president Nancy Birdsall spoke about opportunities and challenges for the implementation of Cash on Delivery Aid, an approach that allows aid agencies to address both short-term and long-term objectives of aid.
Interest in Cash on Delivery Aid has been so strong that we’ve printed a second edition of the book which can be purchased or downloaded online. Here is our new preface:
Since Cash on Delivery: A New Approach to Foreign Aid was published in March 2010, the ideas we proposed have been embraced by presidents and ministers, by heads of public and private institutions, and by researchers and practitioners. The Education Ministry in Malawi sent us a letter asking for help creating a COD Aid program there, the British government has publicly committed to financing pilot experiences, and articles and essays have addressed COD Aid in a range of publications including The Economist, The New York Times, and Public Choice. In the debates that ensued, we have learned even more about the Cash on Delivery Aid (COD Aid) approach and how significant a departure it could be from current aid practices.
One of the first things we learned is just what sets COD Aid apart from other results-based aid programs. While most results-based approaches focus on structuring incentives to change behavior in developing countries, COD Aid aims at changing the behavior of both funders and recipients. Results-based approaches that pay service providers for improving performance, individuals for changing behaviors, or local governments for delivering particular services have their merits and should continue to be explored. However, they are not geared to address constraints to development at the national level and to give recipient countries full flexibility to try interventions or address policy issues outside the domain of the relevant sector ministry. Nor are they meant to make the recipient partner government primarily accountable to its own citizens rather than to the outside donor. COD Aid does all of these things by transferring full ownership and responsibility over strategies to the recipient country.
Feedback on the book also helped us clarify how COD Aid could transform the risks facing developing countries when they receive aid. Currently, aid-dependent countries are vulnerable to changing priorities and domestic politics in funding countries and face considerable uncertainty over how much aid they will receive in any given year. By contrast, COD Aid legally binds funders to pay a fixed amount for each verified unit of progress. A clear enforceable contract with independent verification, as proposed in this book, means that the recipient country only assumes risks related to delivering outcomes. These risks are closely related to the country’s own efforts and are more responsive to its own actions – if not in any given year then certainly over the five or more years we recommend for a COD Aid contract.
Discussions about “preconditions” for successful COD Aid agreements strengthened our conviction that the only true preconditions for this new approach are a good measure of progress and a credible way to verify it. We have heard a number of proposals for such preconditions, but none seem particularly compelling. Requiring that recipients submit plans as a precondition would undermine the “hands-off” nature of the COD Aid agreement. It would perpetuate assumptions (despite substantial evidence to the contrary) that joint planning can substitute for country ownership and that donor-sponsored planning, rather than country-driven experimenting, is the key to progress. Similarly, conditioning a contract on adequate financial controls assumes that it is better to control the use of funds by tracking where they go than to control the use of funds by verifying what they yield. Finally, waiting until countries have information systems in place is a recipe for delay when alternative approaches to measuring progress are available. In short, the key features of COD Aid – defining the outcome indicator, the amount of the payment, the means of verification, and requirements for transparency – are the only real preconditions for COD Aid. Any further eligibility conditions are likely to undermine the restructuring of the accountability relationships or to simply delay implementation.
The limited number of preconditions for COD Aid may make it ideal for the so-called fragile states, countries like Liberia after emerging from civil war or like Malawi after deposing its long-lived dictator. In some of these countries, strong positive leadership emerges but in a context of weak public institutions. Budget support mechanisms cannot be applied fully because recognizable public expenditure frameworks are lacking; and traditional aid projects bypass rather than strengthen public institutions. In such places, COD Aid might be ideal because it effectively controls the use of funds by verifying the progress it achieves rather than the inputs it buys. By working through the government, COD Aid arrangements strengthen public institutions and motivate politicians (and not just technocrats) to care about measuring the country’s progress against clear goals. In these ways, it could help generate the very change that we call development.
Finally, we came to see that the amount funders should offer to pay for each increment of progress is not necessarily linked to the input costs of achieving those gains, which are in any event difficult to assess ex ante in any particular country or setting. In principle the amount should instead be based on how much funders value those outcomes. At the same time, funders justifiably want to get as much value as possible for limited aid budgets, and will also want to avoid overpaying relative to the true costs. If the COD aid payments are comparable with the cost of progress through conventional aid, then they represent good value for money. To the extent they supplement conventional aid – providing an additional incentive for countries to use existing resources more efficiently or triggering helpful changes in political and bureaucratic arrangements – COD Aid payments can be lower than conventional aid that finances inputs at cost. Of course ultimately no one knows how much it costs to alter institutions, reconfigure political bargains or expand capacity in each service in each country. Learning about it takes time and insight about local politics and institutions that can happen when funders focus less on inputs and more on outcomes as envisioned with COD Aid.
In essence, we 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 jointly 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 would change the structure of information reporting and payment triggers for both funders and recipients. Ultimately, it would invite the kind of institution building at the state level that is key to sustainable service delivery and to development itself.
While we were tempted to alter the book and respond to these issues in the main text, we have chosen to leave the text in its current form. The points we have offered in this preface are consistent with and emerge from the principles, analysis, findings, and proposals that you will find here. Readers who take the time to see how COD Aid could be applied to primary schooling will also be able to judge whether we have demonstrated the practicality of the approach. We expect to see a number of COD Aid programs in operation soon. That will be the time to write the next chapter.
Last week UK Secretary of International Development Andrew Mitchell released the outcomes of DFID’s bilateral and multilateral aid reviews.
We were glad to see that the published documents on the bilateral aid review included country summaries that list the funds allocated to each of 27 countries and three regional programs where DFID plans to work in the next 4 years, and the key results these funds are expected to produce. These are likely the highlights of the “results offers” that country and regional teams submitted at the end of last year (as we discussed in this blog).
The focus on results has been a consistent message of DFID under the current government and is reflected in Andrew Mitchell’s commitment to pilot Cash on Delivery Aid this year. All aid programs (of course not just our hoped-for COD pilots) are being asked to focus on results, and allocations are based largely on DFID staff’s estimated costs of the results offers (as explained in the bilateral aid review technical report).
For the multilateral aid review, DFID took a backward-looking approach at the results achieved in recent years by 43 multilateral organizations that it funds, and plans to cut funding to those not demonstrating that they can deliver. To improve its effectiveness, the agency is making tough choices about cutting back on the number of countries it operates in and agencies it works through, despite its overall budget increase.
The two aid reviews are laid out to demonstrate DFID’s accountability and transparency to British taxpayers (“value for money” is heavily emphasized), efforts we very much welcome; we hope that as DFID begins to see the results of its programs – whether good or bad –, these levels of accountability and transparency will maintained, to taxpayers in the UK and to the governments and especially people of aid recipient countries as well.
The maxim that armies are always fighting the last war might just as aptly apply to development agencies: they are too often tackling yesterday’s problems with an outdated set of tools. If our development policies and agencies are to serve our interests, then we need them to both live in the present and prepare for the future. So, what then might development policy look like, say, a decade from now? What should we be thinking about now to get ready? Here are three big trends I think will be shaping the development future:
Drastically fewer poor countries. The old idea of a binary world of a few rich countries surrounded by lots of poor ones (think G8 vs. G77) is simply no longer true. (If you don’t believe me, watch Hans Rosling.) And it’s not just fast-growing India and China, but lots of countries including many in Africa. In fact, IDA, the World Bank’s soft loan window for poor countries and arguably the most important international development institution, will probably lose at least half of its client countries over the next 10-15 years because they will grow too rich to qualify.
New players are breaking the traditional aid cartel. The old donor-recipient model dominated by OECD governments is becoming increasingly passé. New donors (China, India, the Gulf, etc.) play by different rules and with different aid models closely linked to commerce. Big private philanthropists now rival mid-size government donors in both size and influence. The Gates Foundation now distributes about $3 billion per year, which would make it the same size as the median OECD donor (or roughly on par with the aid programs of Italy or Australia). Perhaps most decisively, private sources of capital—hedge funds, sovereign wealth funds, and other investment vehicles—are a growing source of accessible capital for poor regions. Private capital flows into Africa hit $55 billion last year, almost double the level of aid flows.
The likely end of the Gleneagles “doubling aid” era. Recent British and Australian pledges aside, medium-term prospects for major expansions in public sector aid dollars are probably dim. Belt-tightening across Europe and North America will instead likely lead to budget shrinkage and hasten the relevance of private and non-OECD actors in this space.
So, what might all this mean for future development policy?
Escalating demand for value-for-money. This won’t necessarily mean greater enforcement of selectivity or conditionality in the old sense, but rather I suspect we are still at the early front–end of a wave of new results-based aid innovations. Output-based aid is catching on while outcome-based aid (AKA Cash on Delivery) will be piloted soon. We are also certainly going to see new and wider applications of cash transfers, partly because these are proving to be highly efficient and partly as a reaction to old aid models that aren’t.
Global public goods rather than country programs. With only a few dozen really poor countries left, and still large pockets of poor people in middle-income countries, the big value in aid will be investing in solutions that have impacts beyond a single country’s borders. This means investing more in new vaccines, agricultural technology, clean energy, regional infrastructure, and other things that cannot be done by designing narrow country programs for Mali or Cambodia. (Of course political aid and bribes to allies for strategic reasons will stay in a country silo.)
Most importantly, development will shift toward non-aid tools. The debate in the UK right now about its aid program in India is just a sign of things to come. The British public is asking why its scarce aid dollars might go to a country with $300 billion in reserves and a space program. More to the point, India doesn’t want British aid. It wants business, technology, and visas. And this will be increasingly true for places like Indonesia, Nigeria, Mongolia, Vietnam, and Ghana. For the OECD countries, this suggests that development policy will become about fixing the public policies and tools that we have to encourage trade, investment, migration, and other aspects of global interaction that are typically beyond the purview of aid agencies.
Conclusion: We aren’t ready for this new world yet. Unfortunately, western governments haven’t caught up to these emerging trends—and are definitely not set up to deal with these changes. USAID and IDA were established for a very different world more than fifty years ago… and they are showing their age. Most budget systems can’t handle results-based aid models very well, most agencies are stuck in a country-based allocation and strategy mode, and no one yet has a credible or efficient way to fund global public goods. Policy coherence (or even the more modest goal of clear articulation of the tradeoffs) is still more a buzzword than reality. And our private sector tools, like OPIC in the U.S., are woefully underutilized and even under threat.
All of this partly explains why the U.S. is struggling with reorganization and embracing—I believe mistakenly—a flawed whole-of-government model. Creating modern institutions with the capability to respond to this new world and its new demands may sound like moving bureaucratic boxes, but is in fact the next frontier of global development policy. Otherwise, function will continue to follow form, and our development agencies and ideas will continue fighting the last war on poverty rather than the upcoming one.
In a refreshing discussion of COD Aid that’s what Andrew Rogerson calls our idea. Rogerson is an experienced player in aid delivery, having been at DfID and the OECD/Development Assistance Committee. His smart summary covers the latest news, including a pilot of COD Aid for Ethiopia being planned at DfID. It is smartly presented (as in COD implies for the recipient “no free lunch”), with an eye on the practical questions sponsors of COD Aid face.
Here are a couple of issues we are being asked about that Andrew doesn’t cover:
What if a country has no real capacity to measure the outcome to be paid for? Bill Savedoff and I discuss that in the preface to the second edition of our book.
How can a funder do COD Aid in a fragile state? There are different kinds of fragility; COD Aid is worth trying where there is reasonably responsible leadership but weak implementation capacity, as we discuss in our book (pp. 26). Ben Leo reports substantial progress on MDGs in the last two decades in several fragile states with low CPIA ratings (World Bank Country Policy and Institutional Assessment) including Cambodia, Laos, Ethiopia, Malawi, Burkina Faso, and Uganda.
How decide the payment per unit of progress? We think this is the toughest question. The funder has to make an offer but based on what? I don’t think any elaborate studies of unit or marginal costs are needed – but I also think the payment should be below average costs. We are working on this issue and welcome your views.
I recently participated in a conference of Engineers Without Borders Canada which was one of the most thought-provoking events that I’ve attended in a long time. One of the speakers was Vernon Lobo who is with a venture capital firm, Mosaic Venture Partners. For years, I’ve been promoting the idea that aid agencies should behave more like venture capital firms. Like venture capital firms, I’ve argued, aid agencies should not expect every project to be successful. Rather they should think of themselves as having a portfolio of projects – some will succeed and others will fail. The goal for aid agencies is to learn from the whole portfolio so that, over time, good approaches can be retained and shared while bad approaches can be dropped.
What I didn’t expect to hear is how involved venture capital firms are with the firms they invest in.
My experience with aid agencies and multilateral development banks suggests that these well-meaning outsiders are often too involved and demanding toward recipient countries. I’ve come to the opinion that aid is too “hands on” and that country ownership would be better served by a lighter touch. I had thought that venture capital firms were also “hands off” – letting the entrepreneurs they invest in follow their own strategies. To my surprise, Vernon explained that venture capital firms behave very similarly to aid agencies. The relationship with clients is key, so that as unforeseen events occur, the two parties can adapt in good faith with one another. Projects are structured in stages, with opportunities to rethink and revise programs as they evolve. Staging also creates opportunities for the venture capital firm to pull out if they come to the conclusion that further investment is unwise.
While Vernon admitted he knew little about how international aid agencies and NGOs operate, his description of how Venture Capital firms work mirrored the aid agency approach quite closely. This suggests to me that aid agencies are solving a similar problem – risky endeavors with partners – in a similar way to venture capital firms. The one way in which the two differ, though, is quite significant. Venture capital firms do not want to put more money into failing initiatives; while aid agencies do want to fully spend their budgets. This difference is crucial in two ways. Since aid agencies are mostly interested in disbursing, they have weak incentives to find out whether a project is really having an impact or not; the venture capital firms have strong incentives to gather such information. Secondly, the aid agencies are likely to stick with failing approaches much longer than a venture capital firm would.
My takeaway from Vernon’s talk was to question whether my recommendations to aid agencies for more “hands off” approaches are wise. I shared this question with another conference participant who has experience with private and public funders for his NGO. He described how much he valued the support his organization gets from a particular private foundation not just for the financial support but for the skilled and experienced program officers who help him solve problems and improve his operations. He contrasted this experience with the support he receives from a particular public funding agency; in this relationship, he finds the exchanges tend to be more burdensome than helpful.
This left me with one final question. Is this contrast between private and public support a systematic difference? Is there something about public aid agencies and multilateral development banks that necessarily leads them to be more meddlesome than helpful? I know plenty of highly qualified and experienced people in development agencies who are providing excellent support to developing countries. But I also know they are regularly frustrated by the constraints imposed on them to disburse funds independently of progress and to report back to managers in a way that will satisfy political overseers. This suggests to me that the appropriate choice between hands on and hands off is still an open question for aid agencies and one that needs to be tested, as we’ve argued, with something like COD Aid.