We develop a novel approach to measuring long-run economic growth by exploiting systematic variation in the use of color in European paintings. Drawing inspiration from the literature on nighttime lights as a proxy for income, we extract hue, saturation, and brightness from millions of pixels to construct annual indices for Great Britain, Holland, France, Italy, and Germany between 1600 and 1820. These indices track broad trends in existing GDP reconstructions while revealing higherfrequency fluctuations-such as those associated with wars, political instability, and climatic shocks-that traditional series smooth over. Our findings demonstrate that light, decomposed into color and brightness components, provides a credible and independent source of information on early modern economic activity.

Caution: PDF

As a photographer, I’ve noticed that no two photos of a given painting ever look the same. There is much variation due to lighting, color temperature, sensor capabilities, etc. Without controlling for these variables, it’s hard to see how comparisons can be made accurately.

“Our findings [show] that light […] provides a credible and independent source of information on early modern economic activity.”

Wow!

Interesting idea.

I am not 100% convinced by this. The matchup between their painting-based economic index (it's the first component from a PCA analysis, the data for each painting being a vector of pixel-counts for colours in each of 108 bins based on HSV) and GDP growth is pretty dubious, and in places where the two vary together the painting-based metric frequently changes several years before the allegedly-corresponding change in GDP growth.

They have ad hoc explanations for the divergences and try to make lemonade out of the lemons by claiming that their index reveals "higher-frequency fluctuations that traditional series smooth over" but I am willing to bet that if they had had to predict the divergences before doing the calculations they wouldn't have been able to.

I think this is probably mostly pareidolia.

Love it. I think time series views of these things on a site would be fun to watch or put on social media to spread this. Very cool. I appreciate the first citation. I’d vote film as the next medium of interest.

Love it. I respect the first citation.

I would vote to pursue film as the next medium. I would be interested in the predictive potential of your model.

I am not certain this model will teach us a lot but it certainly gets one to think independently which is desperately needed to maintain our humanity.

Thank you for sharing and publishing.

Honestly, I opened the paper pretty skeptical - "colors in old paintings as a growth metric? really?" But the more I read, the more it felt like a clean data-engineering pipeline: HSV extraction, PCA, robustness checks, validation — just with 400-year-old inputs instead of logs.