This report provides detail about the methodological advancements established for improved trade flow modelling at disaggregated level such as HS6. Two approaches are presented and combined. The first one solely splits selected bilateral trade flows by tariff line, while the other disaggregates the GTAP Data Base to more agro-food detail. These methods reduce aggregation biases, especially by depicting substitution effects and the impact of tariff liberalisation more accurately.
Increased detail renders the emergence and disappearance of trade flows more likely, which cannot be modelled based on the commonly applied CET and CES functions. This issue is addressed by defining trade shares of selected products based on a Spatial Price Equilibrium Model. While the standard Armington approach differentiates quality by origin, the Spatial Price Equilibrium Model assumes homogeneous quality for selected products and thus the law of one price. In an example case, all three methods are applied together using the CETA agreement as trade shock.