- The long track record of disasters in Bolivia, Colombia, Ecuador, and Peru highlights their disruptive economic effects, yet for decades, policy responses have been constrained by limited information about their aftermath.
- Nighttime light data offer a new way to track disaster impacts in near real time, providing timely and spatially detailed insights into economic activity—something traditional macroeconomic indicators cannot capture in the immediate aftermath.
- After Hurricane Iota struck Providencia Island in Colombia, a synthetic control approach—comparing actual outcomes with what would have happened in its absence—showed that luminosity fell by 75% that month and remained depressed long after.
When a natural disaster strikes, the world usually sees the same images: rivers swallowing whole neighborhoods, bridges twisted into scrap metal, or families carrying what remains of their homes. What the world does not see is what happens after the cameras leave—those months when local economies don’t just slow down; they flicker, dim, and in some cases, nearly go dark.
Satellite data capturing nighttime luminosity offer a new way to answer a deceptively simple question: what happens to a local economy when a disaster hits and the lights go out? as explored in our latest study, Lights Out: Measuring the Economic Effect of Natural Disasters in Andean Countries.
For decades, the economic story of the aftermath of disasters has lacked information and evidence. Disasters happen in days or hours and have strong effects in the weeks or months following. Yet GDP and other traditional macroeconomic indicators measure activity over months, quarters or years, making them too slow or too aggregated to capture the sharp, sudden economic shock of a flood, a landslide, or a storm. This has made the measure of the impact of disasters tricky.
These challenges can be addressed by looking at something that changes quickly: light, specifically, satellite measurements of nighttime luminosity. Nighttime lights have long been shown to correlate with economic activity. An area that has dimming light may be experiencing negative effects from a shock, such as a disaster. Most importantly, nighttime light data is available for any desired geographic area at a high frequency, allowing researchers to measure the effects of the disaster days, weeks and months after it occurs.
The results are clear. Across Bolivia, Colombia, Ecuador, and Peru, disasters have a negative effect on local economies. Since 2011, these countries have registered more than 1,200 disasters, most of them hydrological: floods, landslides, and storms. These events do not result in high mortality rates (4.6 deaths per million inhabitants), but they disrupt lives on a large scale, with an affectation rate of 825.6 per million inhabitants.
Our results show that in the month a disaster occurs, economic activity (as captured by light adjusted monthly GDP) drops between 0.29 and 1.29%. The recovery process, however, shows mixed results. Although we find a negative effect on economic activity six months following a disaster, this effect is not statistically significant (see Table 1).
These aggregate results, however, make it difficult to truly understand what the impact of a specific event is and how long recovery takes. To shed light on this matter, the paper further analyzes the specific case of Hurricane Iota in Providencia, Colombia, using a synthetic control methodology. This is a way of estimating what would have happened in Providencia had the hurricane not occurred.
In November 2020, Hurricane Iota struck the Colombian island of Providencia with devastating force, damaging 95% of its infrastructure. The images that emerged afterward looked almost apocalyptic: houses without roofs, schools blown apart, the island’s vegetation stripped bare. But satellites captured something even more telling. Providencia’s luminosity dropped by 75% in the very month the hurricane hit.
Recovery was slow. Fourteen months passed before monthly nighttime light returned to pre hurricane levels, and nearly three years before economic activity on the island matched the trajectory it would have followed had the hurricane never struck. Providencia is an extreme case, but the dynamics are familiar across the Andean region, proving that the impact of disasters on local economies is severe.
The study offers a powerful rebuke to the notion that disasters are temporary events with temporary consequences. It shows that even small-scale, frequent events carry a cumulative cost. A flood may last a week. Yet its effects on an economy can stretch for months and years.
Policymakers often speak of “building back better,” but you cannot rebuild what you cannot measure. High-frequency localized economic data should no longer be a luxury. Nighttime lights offer a widely available tool for monitoring disaster impacts as they occur. They allow governments to see which regions need assistance immediately, which are falling behind months later, and which never fully recover without targeted intervention.
The Andean countries (like much of the world) are entering an era where climate shocks are more persistent, more severe, and more economically disruptive. If disasters are going to strike with increasing frequency, then the disaster response must evolve with equal urgency. Waiting for annual GDP numbers is no longer enough, particularly when alternative measures, such as lights, are speaking in real time. Policymakers simply have to look.