Visibility has more to say about the pollution–income link

We show that visibility data contain crucial information for learning the air pollution–income linkage. First, visibility reflects air pollutants (e.g. fine particulates) that are typically omitted by publicly reported indicators of air pollution, typically SO2 and TSP. Second, data on visibility cover almost the entire world, whereas the typical pollution indicators cover a smaller and non-representative sample. We show that both features matter, in a significant way, by employing visibility as a proxy of air quality to re-estimate its relationship with income. Using the findings of Grossman and Krueger (1995) and Harbaugh et al. (2002) as benchmarks, we find that the test of the Environmental Kuznets Curve hypothesis is highly sensitive to the coverage of countries. More importantly, we find that this visibility–income linkage is only partially driven by publicly monitored pollutants, but is dominated by the “unobserved” ones. Addressing both issues, we find the inverse-U shape relationship supported for most of the economies.