Using Neural Networks to Predict Microspatial Economic Growth

We apply deep learning to daytime satellite imagery to predict changes in income and population at high spatial resolution in US data. For grid cells with lateral dimensions of 1.2 km and 2.4 km (where the average US county has dimension of 51.9 km), our model predictions achieve R² values of 0.85 to 0.91 in levels, which far exceed the accuracy of existing models, and 0.32 to 0.46 in decadal changes, which have no counterpart in the literature and are 3–4 times larger than for commonly used nighttime lights. Our network has wide application for analyzing localized shocks.

Here is the AEA-gated published version, and here are other (mostly ungated) versions.  Arman Khachiyan, Anthony Thomas, Huye Zhou, Gordon Hanson, Alex Cloninger, Tajana Rosing and Amit K. Khandelwal, and look for more papers to come in this area.