By Megan Bowen

---

For millennia, humanity has gazed up at the night sky finding meaning in the vast expanse of lights. Now, that trend is being reversed. Instead of looking to the lights in the sky, we’re using satellite technology to look down from above.  

Every night, satellites orbiting Earth capture a glowing network of lights from cities, small towns, and even remote factories. These lights tell a story about human activity.

A recent study by Katarzyna Anna Bilicka, the Lars Peter Hansen Associate Professor of Economics and Statistics at the Jon M. Huntsman School of Business, and André Seidel, associate professor at the Western Norway University of Applied Sciences, suggests this thrum of activity can tell us more than just where people are. It can, in fact, shed light on economic ebbs and flows.

Their work, published as a National Bureau of Economic Research working paper, shows how satellite images can reveal financial misreporting. By analyzing night light emissions from car manufacturing plants worldwide, they found that some corporate financial reports do not match real production levels. In some cases, economic activity is missing entirely from official records.

“Our findings are based on freely available satellite imagery, which comes with a resolution that naturally limits our approach to large-scale industries,” Seidel explains. “That said, commercially available satellite images already offer much higher resolutions, and these continue to improve rapidly. With sufficient funding, near real-time tracking of small-scale global economic activity is maybe already within reach.”

For years, economists have relied on tax records, employment data, and financial statements to measure economic activity. However, these reports depend on companies being honest. That’s not always the case.

Instead of relying solely on company disclosures, researchers are using images from the VIIRS Day-Night Band satellite to track light emissions from car factories and see if light levels can reveal real economic activity.

The COVID-19 pandemic created a natural test for this theory. As factories shut down during lockdowns, researchers noticed a 14% drop in night light emissions. This confirmed their belief that light output is closely linked to production levels.

More importantly, the study found nightlight data explained 70% of variations in factory activities. So, while financial reports can be unclear, satellite images offer a more reliable view of real-world manufacturing.

Missing and Misreported Data

Beyond proving the usefulness of nightlight data, the researchers also uncovered something concerning. Around 50% of firm activities were misreported, and for 38% of turnover, it was unclear where the economic activity actually occurred. In total, nearly $680 billion in global revenue had no clear source.

“What surprised me most is we found half of firm activities were not reported in places where we measured they occurred,” Bilicka says. “This means car manufacturers are moving their reported turnover and profits across firms that belong to them and across different countries, taking advantage of tax rate differences between these locations to lower the amount of taxes they pay in total.”

The study also found a connection between corporate tax rates and financial misreporting. When a country’s corporate tax rate increased by 1%, firms were 20-30% less likely to report financial data. This suggests companies may hide profits to avoid paying taxes, making it harder for governments to track economic value.

This research has major implications for taxation. Many multinational corporations move profits across borders to benefit from lower tax rates, costing governments billions in lost revenue. By using night-light data, policymakers could get a clearer view of actual economic activity. This could help them refine tax policies and ensure companies pay taxes where they truly operate.

“This suggests that when corporate tax rates are high, many firms choose not to report any financial information at all for fear that if they do report it, tax authorities may be able to detect the misreporting,” Bilicka says. “This highlights the difficulties that tax authorities face when they want to verify the information that firms report in their tax returns using financial data. The likelihood of such data not existing increases with the tax rate of the country.”

Although this study focused on car manufacturing, its findings could apply to many other industries, revealing hidden activity and providing real-time updates on economic trends. Bilicka and Seidel suggest future studies should expand their methods to gain deeper insights into global economic activity.

“Looking ahead, I believe methods like ours — when combined with other remote sensing data such as traffic tracking — can provide valuable insights into small-scale economic activity linked to observable physical processes, regardless of where it occurs globally,” Seidel says. “At present, the main constraints on further research are funding and access to high-resolution classical economic activity data, such as daily sales, which are essential for model training.”

As technology advances, satellites could become powerful tools for tracking economic activity and ensuring corporate accountability. One thing is certain: the brightest lights don’t always tell the full story. But with the right data, we can uncover the truth.