By David Poulson
When geographer Joe Messina first analyzed satellite images of Malawi farm fields, he figured he had made a mistake.
Almost everywhere he looked in the East African nation he found maize harvest declines over the previous decade. But this was the site of the Malawi Miracle, a fertilizer subsidy program so successful that it was lauded by researchers in scientific journals and by writers in the New York Times and The Economist.
It became a model program used to justify similar enormous investments by the international community in other African nations.
“I assumed I was wrong,” said Messina, a researcher at Michigan State University’s Global Center for Food Systems Innovation.
And so began a detective story recently published in the journal Nature Plants. It is a story that doesn’t prove Messina wrong. Rather, it reveals a series of missteps, shaky assumptions, faulty data and a desire to confirm success that led other researchers astray.
It shows that the Malawi Miracle fell well short of miraculous.
Not only does the story indicate that one of the poorest countries in the world misspent a significant portion of its scant resources for an agriculture program that didn’t work, it has broad implications for how the international community decides to provide aid on a global scale, Messina said.
It shows that big decisions are made with bad data.
The data reported to the United Nations Food and Agricultural Organization that supported a Malawi success story has problems as mundane as someone incorrectly writing a five instead of a two into a harvest report, Messina said.
Other times, pre-harvest estimates were used to report the actual harvests, he said. Those were off by as much as 30 percent. That’s because the early growing season, when estimates were made, had good rain. But crops were subsequently devastated by severe drought the rest of the season.
And modelling done of the entire nation’s production and that was based on small field samples couldn’t be justified even by assuming the best possible conditions everywhere in the country, Messina said.
Such problems question the assumptions that drive significant and expensive decisions by the international community when it seeks to help hungry countries, Messina said. And it shows what happens when researchers struggle to justify data rather than question when it doesn’t make sense.
“There is a fundamental argument here for the perils of confirmation bias in science,” Messina said.
Messina’s research is controversial. Fertilizer does work in the right places. And other researchers have a deeply held belief in the miracle of Malawi. Messina was shouted at and jeered when he first presented his findings at an Agricultural Innovations conference in Bamako, Mali.
Better data in data-poor regions would certainly improve decisions on how to invest international aid, Messina said. But the larger issue is the complexity of factors that need to be assessed to target aid to where it does the most good.
“They are developing very expensive programs and launching them without consideration of the geographic, the spatial, social, environmental factors that really will limit or promote the success of any of these development programs,” he said.
Listen to a full interview with Messina on this edition of The Food Fix podcast.