Analysis of Panamanian DMSP/OLS nightlights corroborates suspicions of inaccurate fiscal data: A natural experiment examining the accuracy of GDP data

Governments have incentives to misreport their economic productivity to advance their political goals. These incentives have long been understood, but the validity of government data has been difficult to estimate in the absence of viable external estimates. Using historic Defense Meteorological Satellite Program's Operational Linescan System nightlights imagery we corroborate reports that Panama's government data has been increasingly politicised since the handover of the Panama Canal on 31 December 1999. The Canal Handover represents a “natural experiment” in which the production of government data changed in Panama for reasons separate from the desire to manipulate that data. The amount of light a country produces at night, known as nightlight production, has been shown to strongly correlate with GDP. Using subnational Panamanian nightlight production from 1996 to 2012, we detect a significant divergence between the relationship of subnational reported GDP and nightlights before the Canal handover (when the U.S.A. was very involved in their statistical agencies) and the correlation after the handover (with no U.S. involvement). Our results indicate that between 2000 and 2012, Panama reported approximately 19% more GDP than what was expected by their nightlight production from 2000 to 2012, or a total of around 40 billion U.S. dollars. Our results suggest governments may engage in political manipulation of government statistics to improve the appearance of government performance. While indirect data can never definitely confirm economic phenomena, this analysis presents a unique research design and application of historic satellite imagery to corroborate reports of GDP misreporting.