Success looks like nothing happening at all
For thirty years, a quiet network of field workers and surveillance systems has stood between desert locust swarms and the food security of hundreds of millions of people across Africa, the Middle East, and South Asia. A new study has placed a number on that silence: every dollar invested in early warning returns up to 680 in prevented losses, a ratio that speaks to the invisible architecture of catastrophe averted. When conflict in Yemen dismantled that architecture between 2015 and 2019, the consequences were not merely agricultural — they were written into the bodies of 445,000 children who will carry the effects of stunted growth across their lifetimes. The study arrives as a reminder that the most consequential systems are often the ones we never see working.
- A single locust swarm can erase in one day the food that would sustain 625,000 people — a scale of destruction that makes early detection not a convenience but a civilizational necessity.
- The system's greatest weakness is also its greatest proof of value: successful monitoring leaves no trace, making it politically vulnerable to defunding precisely because it works.
- Armed conflict in Yemen between 2015 and 2019 created surveillance gaps that allowed swarms to grow unchecked, cascading into famine conditions that stunted the growth of nearly half a million children — 83 percent of them in countries that never saw direct fighting.
- Researchers used machine learning across three decades of data to make the invisible visible, calculating that stunting from the 2019 outbreak alone threatens nearly 25 billion dollars in annual GDP losses across affected nations.
- As climate change expands locust breeding habitat and intensifies outbreak risk, the international coordination required to maintain these fragile monitoring networks remains underfunded and politically contested.
A desert locust swarm can consume in a single day what would feed 625,000 people, stripping fields and pastures and sending food shortages rippling across entire regions. For three decades, researchers have been measuring what happens when we see them coming — and what happens when we don't.
A study from the National Bureau of Economic Research examined monitoring networks that track locusts across Africa, the Middle East, and South Asia, drawing on thirty years of data analyzed through machine learning. The conclusion is striking: for every dollar spent on surveillance, the system returns up to 680 in prevented losses. The difficulty in demonstrating this value is that effective monitoring is invisible — stopped swarms generate no headlines, no reported damage, no obvious proof of success.
To work around that invisibility, researchers studied what happens when monitoring breaks down. Between 2015 and 2019, armed conflicts across several regions created gaps in field surveillance. Combined with heavy rains that accelerated locust breeding, those gaps allowed swarms to develop and migrate unchecked. The 2019 Yemen outbreak became the defining case study. As civil war disrupted monitoring, swarms spread across the region, and the human consequences extended far beyond destroyed crops. Children exposed in the womb to the resulting famine conditions showed an 18 percent higher rate of stunted growth, and some did not survive to age five. Approximately 445,000 children experienced stunted growth — 83 percent of them in neighboring countries that had seen no direct conflict.
Stunted growth carries costs that compound across a lifetime: reduced cognitive development, diminished earning potential, constrained national productivity. The researchers estimated the long-term GDP impact of stunting from this single outbreak at nearly 25 billion dollars per year — a figure that dwarfs the cost of the monitoring system that could have prevented it.
The study lands at a moment when climate change is expected to expand locust breeding habitat and increase outbreak frequency. The warning networks that currently contain this threat are fragile, dependent on international coordination and sustained funding. The researchers' argument is simple and politically difficult: invest in prevention before disaster strikes, and protect monitoring capacity even when conflict destabilizes the regions where it is needed most.
A swarm of desert locusts can consume in a single day what would feed 625,000 people. When they move across the landscape, they leave behind stripped fields and pastures, triggering food shortages that ripple across entire regions. For three decades, researchers have been trying to measure what happens when we see them coming—and what happens when we don't.
A new study published by the National Bureau of Economic Research examined one of the world's oldest and most consequential disaster warning systems: the monitoring networks that track desert locusts across Africa, the Middle East, and South Asia. The finding is stark. For every dollar spent on surveillance, the system returns up to 680 dollars in prevented losses. That calculation comes from thirty years of data, analyzed using machine learning to trace how locust swarms move from their breeding grounds to the places where people live and farm.
The challenge in proving this value is that successful monitoring is invisible. When field workers spot developing swarms early and intervention teams move in, the locusts are stopped before they cause visible damage. No outbreak means no headlines, no reported losses, no obvious proof that the system works. To get around this problem, the researchers looked at what happens when monitoring fails—specifically, when armed conflicts prevent field workers from surveying locust breeding areas. Between 2015 and 2019, violence in multiple regions created gaps in the warning system. Combined with favorable breeding conditions—heavy rains that locusts need to reproduce—these gaps allowed swarms to develop and migrate unchecked.
The 2019 Yemen outbreak became a case study in cascade failure. As the civil war disrupted locust monitoring, swarms developed and spread across the region. The agricultural destruction was severe, but the human toll extended far beyond lost crops. Children exposed to the famine conditions created by the locusts while still in the womb showed an 18 percent higher rate of stunted growth. Some did not survive to age five. Across the affected region, approximately 445,000 children experienced stunted growth as a result of the monitoring failure—and 83 percent of them lived in countries neighboring Yemen, places that had not experienced direct conflict but were devastated by the ecological consequences.
Stunted growth in childhood carries lifelong costs. It reduces cognitive development, limits earning potential, and constrains a nation's economic productivity. The researchers calculated that the long-term GDP impact of stunting from this single outbreak alone approached 25 billion dollars per year. The monitoring system that could have prevented it costs far less to maintain. The math is unforgiving: prevention is vastly cheaper than the alternative.
The study underscores a broader vulnerability. Climate change is expected to create more favorable conditions for locust breeding in coming decades—warmer temperatures, shifting rainfall patterns, expanding suitable habitat. The systems that currently track and control these pests are fragile, dependent on international coordination and sustained funding. When they work, they work invisibly. When they fail, the consequences are measured in stunted children and lost harvests across multiple countries. The researchers argue for what seems obvious but remains politically difficult: investing in prevention before disaster strikes, and ensuring that monitoring networks can function even when conflict or instability threatens the regions where they operate.
Citas Notables
When locust swarms develop, they destroy crops and pastureland in their path—consuming as much food each day as about 625,000 people.— Eyal Frank, Harris School of Public Policy
A difficulty in assessing those benefits is that when the monitoring succeeds, locust swarms are nipped in the bud and little to no damage occurs.— Anouch Missirian, Toulouse School of Economics
La Conversación del Hearth Otra perspectiva de la historia
Why is it so hard to prove that early warning systems actually work?
Because success looks like nothing happening. A swarm gets spotted and stopped before it reaches farms. No damage, no story. The system is invisible when it's doing its job.
So how did the researchers measure something that didn't happen?
They looked at the opposite—times when monitoring broke down. Armed conflicts in locust breeding areas meant field workers couldn't survey. That's when you see what prevention was actually preventing.
And that's where the 2019 Yemen outbreak comes in.
Exactly. The civil war created a gap in surveillance. Locusts bred unchecked. By the time swarms reached neighboring countries, hundreds of thousands of children had already been exposed to famine conditions in the womb.
The stunting rates—18 percent higher—that's a permanent effect?
Yes. It affects brain development, earning potential, everything downstream. And it's not just individual suffering. When you scale that across 445,000 children, the economic loss to those countries compounds for decades.
So the 680-to-1 return is just from preventing stunting in children?
Just from that. There's also the direct agricultural benefit—the crops that don't get destroyed, the pastureland that stays productive. The real return is probably much higher.
What happens if climate change makes locust outbreaks more common?
Then these monitoring systems become even more critical. But they're fragile. They depend on international cooperation, sustained funding, and the ability to work even when countries are in conflict. That's the harder problem to solve.