Table 1. Experimental design.
Descriptions of the reference (E0), alternative scenarios including trade modeling approach scenarios (E1 – E3) and a scenario testing trade parameter uncertainty (E4).
Scenario | Experiment | Description |
---|---|---|
Reference | E0: Armington with margins and tariffs | The logit based Armington trade modeling approach with calibrated parameters and international transport cost margins and tariffs are introduced into data and model. This experiment provides the updated reference projections to 2100 (5-year step). |
E1: Armington with no margins and tariffs | From the reference scenario (E0), margins and tariffs are removed in data and model. | |
Alternative scenarios: Different trade modeling approach | E2: HOV segmented markets | From E1, large logit parameters (30) are used in Armington to effectively simulate the HOV approach with segmented agricultural markets. |
E3: HOV integrated markets | From E2, regional agricultural productions are pushed to the international market in data and model to effectively simulate the HOV approach with fully integrated global agricultural markets. | |
Parameter uncertainty | E4: E0 with random draws on trade parameter | From the reference scenario (E0), trade parameters are defined as distributions and Monte Carlo simulations are conducted with 1000 random draws from regionally independent parameter distributions. |