Revisiting the environmental bias of trade policies based on an environmentally extended GTAP MRIO Data Base

Recently, Shapiro (2021) identified a new fact which shows an environmental bias of trade policies. This fact reveals that in most countries both import tariffs and non-tariff barriers are much lower on dirty than on clean goods—where “dirtiness” is defined as a commodity’s emission intensity measured in terms of CO2 emissions from fossil fuel combustion per dollar of that commodity’s value. In this paper, we revisit this fact by using a newly-developed environmentally-extended GTAP MRIO Data Base and extend an earlier analysis in several ways. Our preliminary results that rely on the emissions embodied into trade (EEBT) approach and consider import tariffs only (i.e. do not include NTBs) indicate that when looking at the global average indicators, the environmental regressivity/progressivity of trade policy substantially depends on the scope of emissions coverage. We find substantial evidence of environmental bias of trade policies when only CO2 emissions from fossil fuel combustion are considered. In contrast, when a broader set of GHG emissions is analyzed, an opposite relation is observed with more GHG-intensive commodities facing higher tariffs. The latter is largely driven by the fact that import tariffs are high for agricultural and food commodities (e.g. meat and rice) that are relatively clean as they combust less fossil-fuel related CO2 emissions, but are much more emission intensive, when non-CO2 GHGs are accounted for. Our preliminary results also indicate that in the case of fine particulate matter (PM2.5) emissions, one of the major cause of pre-mature mortality worldwide, a positive relation between import tariffs and emission intensities is observed.