04 August 2010

COD aid or cod aid?

So I really want to be enthusiastic about COD (cash on delivery) aid. The idea has been developed by the Centre for Global Development.

COD aid "builds on existing initiatives that strive to disburse aid against results, but it takes the idea further by linking payments more directly to a single specific outcome; giving the recipient country full authority to achieve progress however it sees fit and without interference of any kind from donors; and assuring that the recipient country’s progress is transparent and visible to its own citizens. These features could rebalance accountability, reduce transaction costs, and encourage local innovation and learning."

There is a nice frequently asked questions (FAQ) section on the CGD website that addresses 33 questions.

So why do I have my doubts? My main concern is whether there will be demand for it.

1. Which types of countries will be excited by it?

Countries that are "able but not willing" to prioritise human development are surely not going to be incentivised to become more interested by these relatively small amounts of aid.

For countries that are "willing but not able", where are they going to get the resources from to achieve the outcomes to get the cash?

The FAQ relating to this argues that countries are not starting from blank slates--they have the resources--and COD aid is intended to make sure they make the most of those existing resources. But I think this answer applies best to the "willing AND able" country category.

2. Which countries will want to take the risk of not achieving the outcomes?

Countries cannot have complete control over outcomes. The FAQ response to this states that COD aid is not "all or nothing" and that "a country would be paid a set amount for each additional student that completes primary schooling and takes an approved test, regardless of the level of coverage it reached". Completing school and taking a test seems more like an output than an outcome. The outcome should relate to learning and cognitive achievement. But what happens if the student does not do well in the test? The state cannot be held fully responsible for that, surely.

And are other development outcomes as incremental as education?

It would be good to know if any market research has been done on the demand for COD aid. Obviously demand will depend on the alternatives on offer.

I have not thought nearly as hard about this as the folks at CGD have, and they are smart, so I'm bracing myself for a slew of emails from them.

And one other thing, they really should get another name for this approach. Cod is also slang for "sham". And despite my doubts, COD aid is something to be taken seriously and is definitely not sham aid.

8 comments:

Duncan Green said...

Hi Lawrence, interesting post. I became more pro-COD when I realized that what CGD is actually proposing is really a bonus system. Say 80% up front as normal aid, then 20% extra paid by results to try and improve the alignment of incentives. Worth a try, surely? Check out http://www.oxfamblogs.org/fp2p/?p=2563 for my look at the CGD proposal
Duncan

Paul Isenman said...

Lawrence raised good questions about COD aid. There are also important questions to consider about the cross-country of COD aid if it were to be implemented on the large-scale cross-country basis that the CGD website recommends. My colleagues Geraldine Baudienville and Cecilie Wathne recently considered these questions in a recent ODI paper, ‘Global Funds and Aid Effectiveness’ http://www.odi.org.uk/resources/details.asp?id=4947&title=global-funds-allocation-strategies-aid-effectiveness. Essentially COD aid is a special case of results-based aid distinguished by: a constant payment per specified output across countries; operating only through partner country governments; a highly non-interventionist approach that encourages innovation is how results are reached; and financing for a six-year renewable period, thus easing, or at least postponing, issues of institutional sustainability faced by other versions of results-based aid. Another positive point is recognising the need for indicators of quality, in order to mitigate the risk of letting quantity drive out quality.)

If the principle of donor choice of priorities and per service payments were applied on a large scale across countries, as is intended (according to the CGD website), there would be serious risk of making allocations of aid across countries less effective and equitable, as well as of pre-empting partner country priorities within and across sectors. This would mean a move far from the Paris/Accra objective of ‘putting the country in the driver’s seat’, because donors would be choosing the destinations. The risk of distortion increases with: the number of COD (sub)-sectors covered; COD’s share in relevant (sub)-sectors; and the amount the country receives relative to the cost of providing it -- net of other aid that directly or indirectly reduces that cost. CGD is aware of this risk and calls for COD aid to be additional. With fiscal pressure on aid budgets, though, and with calls for aid for climate change to be additional as well, this does not seem a likely outcome.

But the COD approach remains quite interesting as an additional aid modality if the constant payment across countries is relaxed. And it could be improved if it borrowed from the experience of GAVI and provide up-front support for capacity and systems development for an initial period.

Richard Manning said...

The wish to relate funding to the achievement of results has a long
history, but consistent application of such principles at scale has been rare. Initiatives such as 'Output-based aid' have had some traction, but
have not as yet proved transformative. It is welcome that CGD has been re-invigorating discussion of such ideas. It is a pity that the advent of the Millennium Challenge Corporation was not used to encourage a real test of such approaches, which would have fitted well with its tough criteria for entry, and would have adressed the problem of securing reasonable speed of disbursement.

Account should also be taken of the experience of the 'Millennium Contracts' which the European Commission is using on a significant scale, where a
portion of budget support is conditioned on achievment of MDG-related outcomes. Both aim to avoid the 'all-or-nothing' approach of some forms of
conditioning on results, which can be de-stabilising for recipient
institutions that depend heavily on aid. This is done by the EC by applying the conditioning to only a modest part of its budget support and in the CGD
proposal by paying for incremental achievements which themselves cost the recipient institution money (such as extra pupils at school). Outcome conditionality has some advantages over more traditional process conditionality ('privatise your grain market and we will pay you'), in that it does require observable real-world changes. But the risks are significant, since the recipient institution will seldom be the sole arbiter
of whether an outcome is reached.

These risks loom larger in more fragile environments, where the operating conditions and the ability to access basic data are likely to be problematic. Hence, as with the EC approach, the CGD model appears to have more chance of application in more stable environments - such as those of
recipients of the MCC.

lawrence haddad said...

Thanks for these 3 comments...

Paul Isenman and Duncan Green say that the risks of COD diminish if applied to additional aid.
Richard Manning points out the difficulty of COD in fragile contexts.

Additionality gets around the cash flow probelm for the "willing but not able" countries, although Richard Manning reminds us that this group of countries is pretty large, as it likely includes fragile contexts.

But I would be more convinced if I could actually see some evidence of demand for it...it can't be a totally supply driven solution.

andrew said...

Duncan is right, Lawrence.

Thanks for starting this excellent set of exchanges, by the way.

Much of the debate stems from the difference between an MDG "prize" or bonus amounting small fraction of the cost of, say, putting an extra child through primary school, and the alternative notion that COD pays for the whole cost (averaged across countries), but only retrospectively.

The full-cost version raises several complexities which have come up in the discussion. The most obvious is where does the pre-financing come from to enable aid-dependent countries to obtain the results that justify COD. There are answers to this, but they are messy or raise further issues.

Another concern is that whatever is the level of the average "tariff" per extra child set across countries, if there is only one tariff allowed, this will be more appealing to low-cost countries than high-cost ones. This pattern usually maps quite closely to countries that have a long way to go yet, versus countries that have made earlier progress but now have to tackle harder-to reach gaps. There is an ethical judgment implied in this choice. If you set different prices for different contexts you raise other distributional concerns.

The idea of a prize is also appealing from another perspective, in that it could be run as a fully transparent competition across countries, each getting a share of a fixed pot each year depending on its share of the overall verified results. Emulation and knowledge-sharing could be promoted. A structure like the education Fast Track Initiative would be needed to manage it presumably.

Obviously, payoffs would be (even) more uncertain, as they depend on the relative progress made by others, but still those starting from low levels could get a built-in advantage. As the prize would be fully disbursed each year, moreover, this would avoid locking up reserves and idle balances at country level to adjust for unexpeted over- and under- performance, which could appeal to private and public donors alike.

Finally I cannot resist a comment on the COD title. I believe the term resonates more with those of us old enough to remember when the postman or UPS van driver collected payment for the parcel on delivery.In the web-based world our card is debited before the parcel leaves the warehouse, or even before it is made. We can subsequently send it back and demand a refund, maybe, but that does not work as a good metaphor for aid for progress!)

Best wishes, Andrew Rogerson

(OECD Secretariat, writing here in a purely personal capacity)

andrew said...

Duncan is right, Lawrence.

Thanks for starting this excellent set of exchanges, by the way.

Much of the debate stems from the difference between an MDG "prize" or bonus amounting small fraction of the cost of, say, putting an extra child through primary school, and the alternative notion that COD pays for the whole cost (averaged across countries), but only retrospectively.

The full-cost version raises several complexities which have come up in the discussion. The most obvious is where does the pre-financing come from to enable aid-dependent countries to obtain the results that justify COD. There are answers to this, but they are messy or raise further issues.

Another concern is that whatever is the level of the average "tariff" per extra child set across countries, if there is only one tariff allowed, this will be more appealing to low-cost countries than high-cost ones. This pattern usually maps quite closely to countries that have a long way to go yet, versus countries that have made earlier progress but now have to tackle harder-to reach gaps. There is an ethical judgment implied in this choice. If you set different prices for different contexts you raise other distributional concerns.

The idea of a prize is also appealing from another perspective, in that it could be run as a fully transparent competition across countries, each getting a share of a fixed pot each year depending on its share of the overall verified results. Emulation and knowledge-sharing could be promoted. A structure like the education Fast Track Initiative would be needed to manage it presumably.

Obviously, payoffs would be (even) more uncertain, as they depend on the relative progress made by others, but still those starting from low levels could get a built-in advantage. As the prize would be fully disbursed each year, moreover, this would avoid locking up reserves and idle balances at country level to adjust for unexpeted over- and under- performance, which could appeal to private and public donors alike.

Finally I cannot resist a comment on the COD title. I believe the term resonates more with those of us old enough to remember when the postman or UPS van driver collected payment for the parcel on delivery.In the web-based world our card is debited before the parcel leaves the warehouse, or even before it is made. We can subsequently send it back and demand a refund, maybe, but that does not work as a good metaphor for aid for progress!)

Best wishes, Andrew Rogerson

(OECD Secretariat, writing here in a purely personal capacity)

andrew said...

Duncan is right, Lawrence.

Thanks for starting this excellent set of exchanges, by the way.

Much of the debate stems from the difference between an MDG "prize" or bonus amounting small fraction of the cost of, say, putting an extra child through primary school, and the alternative notion that COD pays for the whole cost (averaged across countries), but only retrospectively.

The full-cost version raises several complexities which have come up in the discussion. The most obvious is where does the pre-financing come from to enable aid-dependent countries to obtain the results that justify COD. There are answers to this, but they are messy or raise further issues.

Another concern is that whatever is the level of the average "tariff" per extra child set across countries, if there is only one tariff allowed, this will be more appealing to low-cost countries than high-cost ones. This pattern usually maps quite closely to countries that have a long way to go yet, versus countries that have made earlier progress but now have to tackle harder-to reach gaps. There is an ethical judgment implied in this choice. If you set different prices for different contexts you raise other distributional concerns.

The idea of a prize is also appealing from another perspective, in that it could be run as a fully transparent competition across countries, each getting a share of a fixed pot each year depending on its share of the overall verified results. Emulation and knowledge-sharing could be promoted. A structure like the education Fast Track Initiative would be needed to manage it presumably.

Obviously, payoffs would be (even) more uncertain, as they depend on the relative progress made by others, but still those starting from low levels could get a built-in advantage. As the prize would be fully disbursed each year, moreover, this would avoid locking up reserves and idle balances at country level to adjust for unexpeted over- and under- performance, which could appeal to private and public donors alike.

Finally I cannot resist a comment on the COD title. I believe the term resonates more with those of us old enough to remember when the postman or UPS van driver collected payment for the parcel on delivery.In the web-based world our card is debited before the parcel leaves the warehouse, or even before it is made. We can subsequently send it back and demand a refund, maybe, but that does not work as a good metaphor for aid for progress!)

Best wishes, Andrew Rogerson

(OECD Secretariat, writing here in a purely personal capacity)

Villette said...

COD aids have their pros and cons, just like anything else. It's not fair to dismiss it completely without seriously testing it out.

Greets feed in tariff consultant