Nighttime satellite images of Earth may provide nuanced measurements of economic growth. According to a new working paper by Professors of Economics J. Vernon Henderson and David Weil and graduate student Adam Storeygard, analyzing changes in an area's "night lights" could be a new means of measuring gross domestic product.
Gathering accurate data on economic growth is often difficult. Moreover, there tend to be gaps in the data — especially in sub-Saharan Africa and developing countries. The authors referred to the Penn World Tables — a standard collection of data on income — to look into flaws in current economic data. Industrialized nations, for instance, are almost always given the highest ranking, while some developing countries are given much lower rankings with significant margins of error. Countries such as Iraq, Myanmar, Somalia and Liberia do not even appear in the rankings.
The Brown economists' work gives an alternative to these measurements. Henderson, Storeygard and Weil looked at changes in light density in U.S. Air Force weather satellite pictures to find growth trends, among others, over a 10-year period.
"As income rises, so does light usage per person, in both consumption activities and many investment activities," they wrote.
For some developing countries, the differences between these estimates were sharp. The data based on the lights implied a 2.4 percent annual growth rate in the Democratic Republic of the Congo, while official estimates pointed to a negative 2.6 percent growth rate.
On the other side, Myanmar's official growth rate is 8.6 percent, but the lights only indicate a 3.4 percent growth rate. The paper also suggests these satellite-based readings account for factors such as civil wars and stagnant markets.
The economists also looked at the relationship between agricultural productivity and increased urban incomes. By studying the effects of "productivity shocks" such as rainfall in 541 African cities over nine years, the authors concluded that increased agricultural output does substantially affect city economies.
The authors do not intend for the "night light" data to replace traditional means of measuring GDP. However, they believe adding data to preexisting estimates will result in more accurate indicators of GDP.