Nighttime Lights and Economic Growth Revisited
Year: 2026
Type: article
Abstract: Gross domestic product is often poorly measured where statistical systems are weak, which has led researchers to seek alternative data whose errors are independent of those in national accounts. Satellite-recorded nighttime lights have become a widely used proxy for economic activity. The foundational framework in the literature treats both lights and official GDP as noisy measures of unobserved true income, combining them into a composite estimator that improves growth estimates where national accounts are poor. Most of the previous literature built on this framework relies on the DMSP satellite, whose data series ended in 2013. This thesis asks whether the relationship persists when the framework is applied to the more recent VIIRS satellite data, a technological improvement over DMSP that economic research has been slow to adopt. The full analysis is carried out for 195 countries over 2012 to 2024: the panel fixed effects regressions, the long-difference specifications, the structural measurement-error model, and the composite estimator. The data-quality classification on which the estimator depends is updated to a 2019 measure of statistical capacity, and the estimates are compared with the original DMSP results throughout. The lights–growth relationship persists under the new sensor. The two-way fixed effects elasticity is 0.225 for the full sample and 0.250 for low- and middle-income countries, and the structural model recovers an elasticity of lights with respect to true income of essentially unity. The differential measurement-error variance across data-quality groups, on which the composite estimator depends, has largely disappeared under the 2019 classification. This is read as evidence of statistical-capacity convergence, though the framework remains applicable once the classification is updated, still assigning lights a meaningful weight for the weakest statistical systems.
Source: Työväentutkimus Vuosikirja
Author: Leinonen, Lasse
Language: English
Subfield: Global and Planetary Change
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7212091266