Ambient air pollution is a significant global cause of mortality. We conduct multidisciplinary modeling to quantify the relationship between international trade and air pollution-related deaths. Our study employs two multi-regional input-output models: GTAP and EORA, which encompass economic transactions across 26 and 65 sectors in 188 and 140 country-based regions, respectively, coupled with a global pollution emissions inventory (CEDS) with 10 anthropogenic sectors. We then use the Intervention Model for Air Pollution (InMAP) to estimate global PM2.5 concentrations resulting from these emissions, and the Global Exposure Mortality Model (GEMM) to estimate excess mortalities associated with PM2.5 exposure. Finally, we measure the externality costs by multiplying the number of deaths by the Value of Statistical Life (VSL) to provide insights for policy and decision-making.
We estimate 3.45–3.55 million deaths attributed to PM2.5 exposure from economic trade activity in 2017, out of a total of 5.2 million deaths due to air pollution. We find that 14–19% of these annual deaths due to air pollution are attributable to international trade across a GDP gap, where higher-GDP countries consume goods or services that lead to air pollution exposure in countries with at least 50% lower per-capita GDP. Generally, high-GDP countries export a substantial portion of their pollution footprint to low-GDP countries, while low-GDP countries import a significant fraction of their pollution burden from high-GDP countries. These mortality exports are complicated by standard methods for assigning economic value to them, which either disregard deaths outside a country's borders or minimize them by using average national income as a scaling factor to apply different VSL values to residents of different countries. We discuss the relationship between this trade in deaths across GDP gaps with commonly-used methods for accounting for air pollution externalities and propose a new valuation method that may help reduce the extent of this outcome.