Urban pollution: A global perspective

We use worldwide gridded satellite data to analyse how population size and density affect urban PM2.5 pollution. We find that more populated and denser grid cells are more exposed to pollution. However, across urban areas, exposure increases with cities' population size but decreases with density. Moreover, the population effect is driven mostly by population commuting to core cities rather than the core city population itself. We analyse heterogeneity by geography and income levels. A counterfactual simulation shows that PM2.5 exposure could fall by up to 40% if population size were equalized across all cities within countries, but the relocation of population from large to small cities that maximizes welfare would be small.