Ideas to Action:

Independent research for global prosperity


Views from the Center


Rakesh Rajani, is founder and head of Twaweza, an initiative that promotes transparency and accountability in Tanzania and other countries in East Africa. This post is based upon comments he made in response to Nancy Birdsall's presentation (see blog post and slides) at the UK Department for International Development on March 9, 2011.

Here are five reasons why I am a fan of Cash on Delivery (COD) Aid:

1. It focuses on outcomes, not inputs. So, for example, what matters is what children are learning, not how many classrooms are built or teachers are trained. Or what matters is how many people are accessing clean water within 500m, rather than the number of wells built.

2. It’s a smarter way of doing accountability. Instead of elaborate governance and checks and balance mechanisms and cumbersome donor involvement, which is a drag for everyone and rarely works well, this approach lays out the outcome and donors get out of the way of the steps in between.

3. It realigns incentives better, to reward what matters, rather than at present where things that are tangential or sometimes even harmful get rewarded (for example, see Lant Pritchett on measurement and development mimicry). With the COD approach, the better you do, the more you get.

4. It is open and transparent, designed to work in a way that everyone can know the rules of the game and follow what's going on, who is doing well and who isn't. As Nancy emphasizes, it enables governments to be more accountable to its own citizens.

5. It is a refreshing antidote to cynicism, not only in places like the UK, where citizens rightly should ask how well the money is used, but places like Tanzania, Kenya and Uganda, where many citizens see aid as a gravy train for those who are in on the game.

Here are five considerations for how the promise of COD Aid could be realized more sharply:

1. Since the whole point of COD is to align systems to a certain defined outcome, setting the right outcome is really important. For example, in education, the goal has been focused on enrollment and years of or completion of schooling, but we know from the work of Eric Hanushek and others that what really matters in education is learning outcomes, particularly early grade literacy, rather than these measures (see also recent research from the World Bank here and here). I worry about some of the example measures given in the COD book, such as counting children who complete school or sit for examinations.

2. Independent verification of the credibility of a government assessment, such as learning achievement of children, is easy to say, very hard to do in practice. Those of you who have worked on audits of government financial systems or performance in relation to other things know this. One can also fudge assessment measures such as terminal examinations, dumb them down, change the denominators, etc. So while it may be heresy in terms of getting behind government systems, I think COD approaches should seriously consider instituting or employing a truly independent (of government) measure of progress/ outcomes (such as or ASER in India Uwezo in East Africa).  I think this is important in any case, regardless of COD, because the risk of conflicts of interest prevailing and depriving systems of truly legitimate measures are simply too high.

3. The COD approach favors cash incentives over layers of technical support and other technocratic interventions.  I agree with this.  But for this to work the incentive needs to be set at the right level. Because CGD is a global level entity, it naturally focuses on incentives for central governments. But with a few exceptions, and in Africa, Ethiopia and Rwanda may be those exceptions, why should we think that central government actors -- ministers of education or presidents -- care much about this approach?  They know that it will take hard work to deliver the results, and that their entire systems are wired differently, such that an additional 20 or even 50 million dollars a year predicated on progress with no fudging doesn't look too alluring.  Perhaps that is one of the reasons there hasn't been a rush of takers of COD Aid on the government side. However, if you turn it around to more local levels, and talk about teachers, and head teachers, local officials, possibly even parents, sharing a pot like $40-50 per child, then this is a considerable sum. Moreover, it is precisely at this level that you have the most leverage to improve learning outcomes; while lots of things can contribute, the single most important factor is likely to be the teacher. The COD promise can really come to life at this local level, and that is why in East Africa we are designing experiments to do COD at local levels.

4. COD has been pitched as a tool to improve public systems, but there is no reason it shouldn't be considered as a tool to incentivize private actors to provide public services, such as secondary education or district level health care.  In East Africa this is already happening through religious organizations and private entrepreneurs.

5. This brings me to my final point. COD has been primarily pitched as a different way of doing aid, which is certainly an important angle.  But I think the real point about meddlesome donors and their often off-the-mark advice has been a tad over-made, and however convenient, donor meddling is rarely the core impediment to progress. A government who knows what it wants and can get things done is likely to know how to put donors in their place. So, from where I stand, COD's bigger promise is not as a better aid mechanism, but as a better way for governments to get better value for money, and as a sharp lever to get results and incentivize those who would do well. That is why the more fruitful locus of discussions about COD may be less between aid giver country and recipient, between say DFID and the Government of Tanzania, but more between reformers in ministries of finance or prime minister's offices and local governments, schools and possibly teachers unions.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.