In the centre of a village in the Nampula region of Mozambique a survey worker spins a bottle and begins walking in its direction to collect data on how many children received a life changing anti-malarial medicine. GiveWell openly admits to this methodology’s use in Malaria Consortium’s surveys, why would they volunteer this? It sounds amateurish, it isn't - but GiveWell still flags the specific way it could skew the numbers.
Trust is a difficult thing to earn, most organisations create trust through confidence, brand-power, and results. GiveWell doesn’t just create trust, they provide it, they show their working. By transparently discussing flaws and potential issues in the programs it reviews GiveWell reaches a level of authenticity and honesty that I have not seen in many people let alone organisations that exist in a harsh competitive world. GiveWell provides trust through its uncertainty.
There are multiple places in GiveWell’s review on Malaria Consortium’s seasonal malaria chemoprevention (SMC) alone where they point out places of concern. GiveWell highlights uncertainty over how many children die of malaria in the areas it supports, the lack of audits in most surveys, the 3-4% of audited interviews found incorrect or fabricated, survey results clashing with each other, along with other problems. The survey by Bottle-Spin comes into question itself, GiveWell says it could bias results upward, overstating how many children were reached. When two surveys of the same campaign differed by as much as 24%, it relied on the one it judged more accurate, the one reporting lower coverage. While some of these issues do seem to be impactful, the point is that GiveWell highlights them, discusses them openly, weighs the risk, and then finally shows that even with these issues the main purpose of what they are doing is still achieved, lives are being saved.
GiveWell’s concerns over the success rate of SMC led them consider whether other malaria funders would conduct their own SMC campaigns. While once again this may seem damning this aligns well with why they should in fact pursue this campaign, the area of support is neglected and therefore impact levels should be higher than in other interventions.
