Satellites add a new layer to global poverty data

"On paper, Arcelia looks like a poor-but-average Mexican town. It sits in Guerrero, Mexico's second-poorest state. Official data gives it a score of 0.714 firmly in the "high development" band on the United Nations' Human Development Index (HDI). Then a satellite looks at Arcelia. Using artificial intelligence to analyze what it sees, it returns a lower score of 0.617."

"By the UN's own classification, that is no longer high but medium development a different development tier and a different policy-reality for 33,000 people. Arcelia is not a special case. More than half (58%) of the global population is in the wrong development tier, because official data averages too broadly to see them. That is the central finding of a study published in the journal Nature Communications by researchers from Stanford University in the US and the UN's Development Programme."

""There hasn't been a census in the last 10 years in about half of the world's poorest countries," said Hannah Druckenmiller, a co-author of the study, highlighting the necessity of up-to-date and accurate information to ensure public policy matches people's day-to-day needs. An accurate HDI score matters for aid delivery The Human Development Index is not merely a ranking. It "can determine allocations of global resources," the study's authors note."

"The problem is that the HDI only provides a score for entire countries. It wasn't originally conceived to differentiate between provinces or even municipalities within a country. But in a simulated aid program for Mexico, one that targeted the poorest 10% of the country's population, researchers found that adding data from the municipal-level improved their understanding about the state of people's development levels of poverty and wealth, education and health by more than 11 percentage po"

Arcelia in Guerrero, Mexico, has an official HDI score of 0.714, placing it in the high development band. Satellite imagery analyzed with artificial intelligence produces a lower score of 0.617, which falls into the medium development tier. This tier change implies different policy realities for about 33,000 people. The issue is widespread: 58% of the global population may be in the wrong development tier because official data averages too broadly. Many of the world’s poorest countries lack recent censuses, limiting up-to-date accuracy. HDI scores influence allocations of global resources, so local misclassification can cause aid to miss people most in need. In a simulated Mexico aid program targeting the poorest 10%, adding municipal-level data improved understanding of development and poverty, wealth, education, and health by more than 11 percentage points.

Read at www.dw.com