Bright Lights, Rich Cities

Staff Writer

Fly into a city on a clear night and you’ll enjoy a jeweled panorama below. Starkly lit skyscrapers, yellow glows from the windows of scattered homes, and roads pulsing white and red as cars speed over highways and crawl along residential streets. All of this electricity usage tends to indicate that a city has money to spend. And, indeed, in a new study, economists have shown that satellite measurements of an area’s lights reveal how economically developed it is.

Judging the economic status of a country like the United States is relatively easy: the relevant numbers are carefully reported and widely available, allowing anyone to easily calculate measures of economic development such as the gross domestic product (GDP). Measuring GDP isn’t as easy for countries with poor financial reporting, such as many African nations, or biased reporting, as many claim China has. And economists might as well forget about documenting economic growth patterns from almost any small, remote region.

In the hope of finding an alternative means of measuring GDP, macroeconomists Xi Chen and William Nordhaus of Yale University turned to nighttime images of the globe taken by U.S. Air Force satellites. They overlaid a grid on these high-resolution images and measured the amount of light, or radiance, emanating from each cell. For countries that already provide economic information widely considered trustworthy, brighter radiance in a given region correlated tightly with a higher GDP in that region, the researchers report online today in the Proceedings of the National Academy of Sciences. Using this data, the economists built a statistical model, correcting for certain variables such as location on a continent (reflections on large bodies of water can distort satellite images) and distortion of the image by the atmosphere, that then allowed them to estimate the GDP for countries with unclear economic status.

This proof-of-concept model can be applied to many scenarios for which GDP is unknown or underreported. “It’s been about building a tool; the really exciting data will come when people start using this tool,” says David Weil, an economist at Brown University, who has been independently working on a similar project. Satellite nighttime images alone are a poor measure of GDP, he cautions, but the beauty of the strategy is that it can be added to government reports, even erroneous ones, to increase overall accuracy.

Nordhaus agrees that one has to be cautious when using the nighttime imaging model. For one, cities with excellent economies can differ significantly in their light output. (Las Vegas, with its flashy casinos that run all night, far outshines the more sedate but similarly economically productive Salt Lake City, whose residents tend go to bed earlier.) And the Air Force developed its satellites to analyze cloud cover, not city lights, so the images themselves are not optimized for this kind of study.

Still, the researchers say that the new measure will be useful for innumerable projects, such as understanding the borders of tribal regions in Africa and how their economic growth compares. Nordhaus suggests that this night-light strategy could have been useful prior to the U.S. invasion of Iraq: little about that country’s economic distribution was known at the time. And economists interested in public health, civil conflicts, or environmental changes could use the images to understand how these factors affect a country’s GDP could over time.

Posted in Economics