Nancy Birdsall and William Savedoff.
CGD is following the growing number of development programs that experiment with results-based approaches, many of them influenced by the concept of COD Aid. The design stage is important to a successful COD Aid agreement but a critical moment occurs when the first results are reported. Will the parties honor the agreement as written and uphold it or renegotiate and undermine its integrity? In this blog we discuss this critical moment in terms of finding ways to limit the pressures to disburse precipitously. In our next blog, we will explain how setting targets subtly undermines the focus on actual performance and the ability to stick to an agreement.
An increasing number of aid agencies are experimenting with programs that incorporate the main features of COD Aid: paying for outputs, giving the recipient greater discretion to spend as they see fit, independent verification, and transparency. (See our brief and book for more details). We’ve argued that the design of COD Aid programs can be rather easy, though the quality of the indicators chosen and the verification process are certainly critical to success. We have spent less time talking about what happens once the program is up and running. In particular, what happens when you find out how much progress actually occurred?
One of the seven sins of the official aid system is impatience – impatience for results as well as impatience to disburse. When development projects go well, this impatience is fine. The project succeeds, the money is transferred, and everyone goes home satisfied. However, development is not easy or predictable and, most of the time, pressures to disburse run into a brick wall – the results were not forthcoming.
Agencies find ways to manage the fallout from such collisions by re-interpreting results and extending waivers, especially when results are complex, multifaceted, vague or simply not measured. But one of the advantages of results-based programs that follow concepts we have described in Cash on Delivery Aid (COD Aid) is precisely its explicit link between funding and one or a few simple outcome indicators. In other words, a well-designed COD Aid project reduces ambiguity and at the end of the first year – when the first progress report is submitted – aid agencies and recipients alike are forced to confront the gap between the results and disbursements they expected with whatever happened in reality.
At this point, whether the surprise is positive or negative, pressures will necessarily arise to renegotiate the agreement. This is why agencies need to plan ahead when they tie project disbursements to results. They have to have contingency plans for underspending as well as overspending.
Don’t get us wrong. Impatience to disburse is usually well-intentioned. People who work in the system are well-meaning, dedicated, can-do types, with a mission: to get the roads built, the schools functioning, and the fertilizer delivered as quickly as possible. Meanwhile they need to tell their senior bureaucrats at headquarters in Berlin, Tokyo or Washington what inputs they will be purchasing at what cost for each of the next five to seven years. This way the input costs can be included in the budget request that goes to the donors’ legislatures. Aid agencies also are sensitive to the impact of financial flows on recipient governments, whose education, transport and finance officials want to include the expected disbursements (grant income) in their forward-looking planning and budgets.
All this only works, however, if development is about buying and applying inputs in a world with substantial certainty. In fact, development involves a lot more. Even the simplest infrastructure projects require the development (yes, it may sound redundant but that’s why we use the “D” word) of institutions, organizations, and practices that assure proper maintenance and, we hope, capacity to plan and execute new investments. And because development involves so much more, the probability that plans go awry, that delays occur, that results fail to materialize, is quite large. Essentially, development projects need better ways to manage this uncertainty when planning how much and whether they will disburse.
In designing COD Aid, we anticipated this problem and suggested a few ways to manage this impatience to disburse (see, for example, our FAQs, this workshop summary and a paper by Ben Leo). We’re not the only ones to write about this issue, as can be seen from the OECD’s take on mitigating the unpredictability of aid spending. The first best strategies are to tackle whatever bureaucratic constraints are in the way, such as requiring money to be spent in the year it is budgeted. When this isn’t possible other strategies are available and generally fall into three complementary categories: realistic expectations, diversification, and contingency plans.
- Realistic expectations: Recognize the overly optimistic bias of development programs and stop believing that the future is significantly different than the past. It is common to introduce contingencies in budgets – an additional 10 or 20 percent in anticipation that costs will be higher than expected. Why can’t disbursement plans make room in the other direction, reducing planned expenditures by 10 or 20 percent per year (or more) in the early years and extending payment timeframes longer?
- Diversification: The variability of individual project performance is likely to be higher than the variability of a group of projects combined. This is particularly true when they start at different times, take place in different countries, and are in different sectors. Taking advantage of this, disbursement plans for an aid agency’s overall budget can be managed by planning for disbursements on the portfolio of projects rather than insisting that each project meet its own disbursement plan. Recipient governments can do the same, focusing on budget planning envelopes with regard to a group of aid programs.
- Contingency Plans: If funds set aside for a particular project or country have to be spent at the end of the fiscal year and underperformance is a possibility, then establish a clear contingency plan. Can the project or country-specific funds be held in trust for a year (perhaps as no more than an accounting device, since a donor’s budget is fungible and some other project may have over-performed)? What other related forms of spending would complement the development project without compromising its success? A results-based project could stipulate that if a specific program doesn’t deliver the expected results, the remaining money will be transferred to a proven and highly regarded international program, such as an international vaccine program, a research pool, or a climate change mitigation fund.
There are certainly other approaches and the best way to proceed will depend on the particular features of an agency’s financial management. The key, like any effort to change one’s habits, is to face the problem and plan to do something differently in the future with a clear commitment device. When designing results-based programs, like COD Aid, this is particularly important so that the surprise that comes at the end of the first year is … well … expected.