First author coreleates night light with GDP growth, shows example of India. They he questions countries like China, North Korea and Russia. He shows the satellite image of North Korea.
Where is equivalent data for China and Russia ?
The amount of street lighting is a complex function (varying a great deal in space and time) of local budgets, changes to CFL/ led/ other technologies, and political desires on the one hand for the perception of fighting crime, and on the one hand much more rarely for less light pollution. I do not see that they controlled for any of this (and can't imagine how they could have). Possibly some of the cited papers that I did not look at might have tried to address this.
I would suggest an alternate interpretation: China (and maybe Russia as well) don’t actually know what their GDP is either.
While beginning reading I was ready to dismiss this as a correlation/causation sort of thing.. but found myself pleasantly intrigued with the analysis.