n.V3-3.06 | IMPACT EVAPORATION
01| Money goes into a poor community. A programme runs for three to five years. Schools get built, wells get dug, farmers learn techniques, health workers get trained. The endline assessment records success. The organisation leaves. And then, quietly, the gains start to come apart.
02| The sector calls this the sustainability gap. A more honest name is impact evaporation.
03| The mechanics are not mysterious. The standard endline evaluation is run in the final weeks of a project, while staff are still on the ground and participants are still engaged. It measures the moment of maximum effort. It does not measure the moment of truth, which arrives eighteen months after the last vehicle leaves.
04| WALA is the case that should be taught in every development programme. The Wellness and Agriculture for Life Advancement project ran in southern Malawi from 2009 to 2014, USAID-funded, implemented by Catholic Relief Services. Its endline reported a substantial reduction in child undernutrition and in communities' need for food aid during crises. In 2018, four to five years later, AidData and Mathematica went back:
"In the wider, quantitative sample, there are no differences in participation in agricultural, livestock, environmental, health or food-related activities between WALA and comparison villages."— Ben-Yishay et al. (Long-term Impact Evaluation of the Malawi Wellness and Agriculture for Life Advancement Program, AidData/Mathematica, 2019, p. 4)
05| The qualitative half is worse, because the researchers deliberately selected the villages where WALA had gone best:
"Even in the particularly successful villages selected for the qualitative study, nine out of ten agriculture and environmental management activities were either discontinued or had mixed results for sustainability."— Ben-Yishay et al. (2019, p. 8)
06| WALA is not an outlier chosen for effect. It is one of the very few projects anyone bothered to check. That is the second half of the problem: ex-post evaluation barely exists. Projects are evaluated at closure, when the answer is flattering, and almost never four years later, when it is not. The industry's confidence in its own best practices is therefore drawn from a biased sample — the short-term wins, not the long-term survivors.
07| The most careful multi-country work reaches the same conclusion without the drama. Rogers and Coates studied four food assistance projects specifically to learn what survives exit:
"Evidence of project success at exit did not necessarily imply sustained benefit over time."— Beatrice Lorge Rogers & Jennifer Coates (Sustaining Development: A Synthesis of Results from a Four-Country Study of Sustainability and Exit Strategies among Development Food Assistance Projects, FHI 360/FANTA III, 2015, p. 18)
08| Most projects do have an exit strategy. In practice it is a handover document delivered to a local government with neither the budget nor the staff to honour it. Practitioners have a name for this: checklist sustainability. The box gets ticked, the NGO departs, the well stops working.
09| Career incentives keep it in place. Development professionals are assessed on disbursement and on short-horizon KPIs. Nobody is promoted for a project still running ten years after they moved on. The system rewards the launch, not the legacy.
10| There is a subtler damage that no evaluation records. Deliver a service free for five years and then withdraw, and you have driven out the local water vendors, food suppliers and informal providers who were serving that community before you arrived. When the free version stops working, the market that preceded it is gone too. The community is left with neither. Nobody counts this, because nobody is there to count.
11| And the same error is now being loaded into climate finance. Sea walls built, mangroves planted, early-warning systems installed by visiting organisations. When the organisations leave and the maintenance budgets do not follow, the resilience is gone in the first serious storm season. A sixty-year-old broken model, applied to the most urgent problem of the century.
12| The deepest finding is the one that should worry us specifically, because it targets our own claim rather than theirs. Katherine Casey synthesised randomised trials across community-driven development programmes — the closest thing the sector has to what we propose:
"I synthesize findings from randomized controlled trials and find that CDD effectively delivers public goods and modest economic returns at low cost in difficult environments. There is little evidence, however, that CDD transforms local decision making or empowers the poor in any enduring way."— Katherine Casey (Radical Decentralization: Does Community-Driven Development Work?, Annual Review of Economics, 2018, Vol. 10, p. 139)
13| Read that again as a warning aimed at IkoCiti. Community participation delivers goods. It has repeatedly failed to deliver durable power. Casey's explanation is that the institutional arrangements a programme introduces are not adopted by the community for anything else, and do not change the traditional institutions they sit beside. The programme's committee meets while the programme is funded. Then it stops.
14| This is not a funding crisis. Decades of money have been spent — the OECD's own aid series runs to several trillion dollars since 1960, and private philanthropy and remittances are larger still. It is a design crisis, rooted in one assumption: that change imposed from outside, measured on the outside's timeline, by the outside's metrics, will take root and hold once the outside leaves.
15| It rarely does. And a platform that intends to claim otherwise had better be able to say exactly what it does differently, and then prove it with numbers it did not choose in advance.
