This study estimates month-specific yield associations between U.S. corn yields
and five air pollutants: fine particulate matter (PM2.5), ground-level ozone at 60
and 80 ppb thresholds (AOT60 and AOT80, respectively), nitrogen dioxide (NO2),
sulfur dioxide (SO2), and carbon monoxide (CO). We draw on data that combine an
extensive commercial variety trial dataset of 210,207 observations from 17 U.S. states
(2000-2023) with EPA Air Quality System monitor measurements. Our findings show
that the aggregate effect of pollutants during the growing season is mostly negative
yet minor or statistically insignificant; however, monthly decomposition uncovers
systematic heterogeneity. For PM2.5, March exposure is associated with a 1.3 bu/acre
yield reduction per μg/m3, while August exposure is associated with a 1.0 bu/acre
increase, a seasonal pattern consistent with an aerosol scattering channel. The August
AOT60 coefficient is −0.030 bu/acre per ppb·hour, with a more pronounced negative
effect under AOT80. NO2 shows negative associations in March and August; SO2 and
CO show none. Interestingly, the estimates provide little evidence that genetically
engineered varieties mitigate yield losses from air pollutants; if anything, they suggest
larger losses, except for the PM2.5 interaction.
