Hand a whole village enough money to double what it spends, and the textbook says the shopkeepers will put their prices up. Inflation eats the cash and everyone finishes where they started. GiveDirectly ran that experiment in a Malawian subdistrict called Khongoni, where the injection was worth about as much as the whole local economy. Prices went up roughly 1%, then fell back to nothing inside five months.
The design is what makes it an experiment rather than a donation. Every adult in a treated village gets the money, unconditionally, in two payments: about $50 shortly after signing up and about $500 roughly two months later. Some villages get it and others do not, and the treated ones are grouped so whole areas end up heavily covered while neighbours are barely touched. That variation is the instrument. It lets researchers watch a market where most customers suddenly have money, against one where only a few do.
What the shopkeepers actually did
Three things kept prices down, on GiveDirectly's account. Households spread the money over months rather than spending it at once. Shoppers had somewhere else to go, including the markets in Lilongwe, so a trader who marked up lost the sale. And traders said outright that gouging would cost them their reputation once the money was gone. Where demand did jump, vendors mostly ordered more stock, and new traders turned up to sell it.
A year after the transfer, the share of adults in Khongoni living above the extreme poverty line, $3 a day, had roughly doubled, and household spending was still 89% higher than before. About a third went into housing, livestock and solar panels rather than food, the sort of buying that keeps paying after the money runs out. GiveDirectly is careful about that doubling: its own estimates of the starting share range from 18% to 36%, and it says the truth sits somewhere inside that.
The district version, and what it is testing
Chiradzulu is a different size of thing. About 400,000 people live there, and GiveDirectly wants the programme reaching 185,000 of them by early 2027.
The money is mostly Canva's. The design company's foundation committed $100 million to the scale-up, the largest single gift GiveDirectly has had.
The businesses are studied too, which is why it had to be done at district scale. About 2,000 local enterprises get a cash grant plus advance warning that demand is coming, another 1,000 get the warning only, and roughly 7,000 will be surveyed across the district's 77 weekly markets.
GiveDirectly's own benchmark is Kenya, where it measured $2.50 of economic activity for every dollar handed out. Whether Malawi repeats that is what the trial is registered to find out: it sits with the American Economic Association, is led by the Oxford economist Dennis Egger, and runs to 2035.
