Satellite images reveal how Putin’s economy is really performing [premium]

A team of researchers from Georgia, the Czech Republic and Slovakia has published a working paper that uses nocturnal satellite imagery as a proxy for local economic activity across Russia. The paper was posted on CESifo/SSRN in August 2026 and summarized by its authors on CEPR/VoxEU on 19 August 2026. The analysis relies on the publicly available NASA VIIRS Day/Night Band product, commonly known as the ‘Black Marble’.

The authors assemble a panel covering 1,058 Russian cities and analyze monthly radiance values from January 2012 through January 2025 (omitting some summer months at high latitudes). They treat night‑time light intensity as an observable correlate of economic activity and apply difference and discontinuity designs plus regional aggregations to estimate how patterns changed after the start of the full‑scale invasion in 2022.

The paper reports two principal results. First, average night‑light growth slowed after the onset of the war, which the authors interpret as evidence of an economic deceleration that shows up in satellite data. The slowdown is described as persistent rather than a transient dip.

Second, the effects are far from uniform across the country. Western and north‑western parts of Russia exhibit larger contractions in night‑light intensity, whereas southern and eastern border regions show relative increases. The authors suggest these spatial patterns point to a redistribution of activity — potentially reflecting diversion toward non‑Western markets or war‑related activity concentrated in particular regions — rather than a uniform national recovery.

These conclusions are drawn cautiously as statistical patterns, and the researchers explicitly contrast them with Russian official statistics that claim a recovery in 2023–25.

Night‑time light data are a widely used alternative indicator in contexts where official statistics may be incomplete, lagged, or disputed. Light intensity correlates with urbanization, infrastructure use and economic concentration, and the VIIRS product offers high spatial resolution and global coverage.

Yet the authors and the existing literature acknowledge several limitations. Bright cities can saturate sensors, weather and seasonal gaps can affect readings (hence the paper’s omission of some summer months), and increases in lighting do not always equal higher civilian production — they can stem from military installations, temporary works, or changes in public lighting policy. The study deploys methods to mitigate these issues but cannot eliminate all measurement uncertainty.

The study highlights a notable divergence between satellite signals and Russia’s official macroeconomic narrative. While government statistics point to a rebound in 2023–25, the night‑light trends do not align with a nationwide recovery. The authors stop short of accusing statistical agencies of manipulation; rather, they stress that different indicators capture different facets of economic life.

That methodological plurality has practical consequences: analysts and policymakers should not rely solely on headline GDP figures when assessing the depth and geography of economic change in an economy exposed to war, sanctions and large‑scale structural shifts.

If the satellite evidence of a sustained slowdown in western Russian industrial hubs is correct, consequences will ripple through European business and finance. Lower demand in those regions reduces markets for industrial goods, consumer products and services from EU exporters. A spatial shift of activity toward southern and eastern regions would reinforce trends of reorientation away from Western economic ties.

For European policymakers this means sanction effects and trade realignments may produce long‑lasting structural change, while firms need to reassess country and regional risks more granularly. Banks and investors can use remote‑sensing indicators as supplementary early‑warning tools for regional exposure, and governments can combine such tools with traditional statistics to better anticipate impacts on energy markets, commodities and supply chains.

Important uncertainties persist. Night lights cannot fully substitute for direct measures of GDP, employment or investment. Brightness changes may reflect military infrastructure expansion, temporary construction, migration‑driven shifts in urban density, or energy distribution changes rather than pure civilian economic growth or decline.

The dataset used ends in January 2025; subsequent developments could alter the picture. The authors themselves call for further work combining night‑light measures with other high‑frequency indicators — for example, electricity consumption, mobile‑phone activity or trade flows — to disentangle the mechanisms behind observed spatial shifts. For analysts, the central takeaway is methodological: integrate multiple data streams to form a more robust and geographically nuanced assessment of Russia’s economic trajectory.