People in a conflict affected and neglected area of Cameroon get cash transfers by surprise. No field assessments, no selection of households. Random gifts.
-AI summarized my idea for clarity and I rewrote a few points.-
Aid without aid workers" sounds terrible. I know. I have been an aid worker for thirteen years. Let me tell you why I want to try it.
In 2018 and 2019, in Cameroon, I watched people rush to join the aid response because it paid better than anything else in a place with no jobs. A handful of them are dead now. Almost all of them have been kidnapped at least once, even if briefly.
What were they risking their lives for? Mostly to "target" and to run "assessments": surveys, lists, verification. Expensive work, and very often it produced data nobody could use.
Then aid distributions. Cold dangerous places that look horrible on their best day. I know of a child was run over by a truck at a distribution site by the road. A guerrilla fighter who tried to keep order by shooting was killed by the crowd trying to overrun the site. I am not sorry for him, but think about the level of violence it takes to kill a man with bare hands. That is what distribution sites are, no matter the PR we use to spin it.
"Targeting" tries to find the "real poor" the "most vulnerable." In reality it divides communities into the eligible and the not eligible.
A month later, a chunk of the people we left out are more vulnerable than the ones we chose, but changing the list is seen as unprofessional, so we keep widening the gap. Targeting takes longer than planned, so we run on the same list for a year or more, through riots and thousands of complaints.
All of this makes aid more expensive and adds no value, is just keeping people busy, at best.
And even if none of this happened, the question would stand. Aid is sent for the people who need it, not for the workers and organisations that deliver it. If we can reach 20% more people by employing fewer of us, why shouldn't we? The mission of an NGO is to help people, not to employ them. The mission of a donor is to help people who need help, not to sustain operations and payrolls.
So I want to try to do away with it. Choose the place, get random accounts from there. Send cash by mobile money to people, with no surveys, no lists, no distribution sites and no monitoring missions after. The only staff left are the ones the money doesn't need: the paperwork, payments and reporting, which AI can now do, with three humans overseeing every batch.
Some money will reach people who are not the poorest in the village. They will still be among the poorest people in the world, people like the one who left the village recently to struggle in town, or the wealthiest man in the village, who works every day in a farm and has a fridge to sell cold drinks. I am OK with that targeting "error".
The first batch is 40 gifts in Lebialem, one of the worst areas of Cameroon.
Prove the pipeline end to end. Funds in → batch prepared by an agent → two humans approve → MTN pays → agent reconciles bank debit, MTN report and transaction IDs → public report. Target: under 5% failed payments, and over 95% of donated money reaching recipients' wallets.
Publish everything. A public page shows every batch: amount sent, number of recipients, fees, failed payments, thank-you replies, and what recipients said about where they are.
Prepare the random-draw model. In parallel we are asking MTN to run a fully random draw among active subscribers near selected cell towers: MTN keeps the list and shares only its size, we send back random positions, and MTN pays those. No personal data leaves MTN. This first batch builds the track record to make that request credible.
The idea is to prove that the combination of cash, random lotteries in the poorest places on earth, and operational overhead by agents can be a useful addition to the humanitarian and development landscape. Not only costs are saved. Lebialem has had very little assistance since the beginning of the civil war because it is expensive to travel there and has been extremely dangerous for aid workers.
We will give recipients the option of saying thank you through sms or whatsapp with a contact number. We will verify the payment reports from the mobile money company.
All in cash transfers plus the operators fee for transfers and sms.
Marc Serna Rius (approver and accountable lead): 13+ years in humanitarian operations in Jordan, Afghanistan, Cameroon and Nigeria with UN and NGOs. Certified in fraud examination and internal investigations.
Effective Help Association (Buea, Cameroon): registered association since 2023, with a GlobalGiving profile. Runs cost-effectiveness analyses and evaluations for local organisations. Two board members approve every payment batch.
AI agents: prepare payment files, reconciliation, reporting, recipient messaging and fundraising
MTN onboarding is slow or refused for an association. Fallback: Orange Money, or a licensed payment aggregator.
Recipients distrust an unexpected gift and treat it as a scam, or scammers exploit the moment. Mitigation: the SMS asks for nothing and warns that genuine gifts never require a fee.
Safety in a conflict area: sudden money could draw attention. Mitigation: modest amounts, spread across villages and days.
Little feedback: few thank-you replies leaves a thin picture of who was reached. Acceptable risk, we will report it honestly.
Serving the wrong person: We could get the richest person in the village or someone who has travelled to an urban area. We are comfortable with that, this person would still be in the income bottom globally and the costs saved in targeting are worth it. Literally if we had to do this project with traditional community mobilization and surveys we would be spending 3,000 to 4,000 dollars to transfer the remaining 1,000 and we would have a 25-55% chance at a kidnapping.
If it fails, the most likely outcome is a smaller batch or a switch of payment provider, not money lost: unspent funds roll into the next batch, and we report the failure publicly.
We have raised about 8,000 USD for other projects. But not for this idea, this is the first funding platform we approach.
Comments
No comments yet. Sign in to create one!