Research 26 Foregone tariff revenues As 75% of all customs duties collected from the member states are transferred to the EU revenue for customs duties amounts to more than 10% of the EU budget. In 2014, total customs duties in the EU summed up to around EUR 21.9 billion of which EUR 16.5 billion became part of the EU Budget.13 In the case of CETA, the EU Commission (2016) esti- mates that foregone tariff revenues from Canadian imports into the EU are EUR 158 million per year from industrial goods alone. When tariffs on other goods (agricultural and food products) are added, the foregone revenue for the EU would increase to around EUR 330 million per year or EUR 3.3 billion over a implementation period of 10 years (UNCTAD- TRAINS data, average 2010-2014). Even though these results are only a rough calculation, data on potential adjustment costs and foregone public revenue are lacking in CETA assessments. These potential downsides are even more relevant in the case of marginal benefits as reported for CETA. Again, po- tential gains range from EUR 4 billion (SIA, 2011) to EUR 12 billion (Joint Study, 2008) after an implementation period of up to 10 years. In contrast, adjustment costs on EU labor mar- kets and foregone tariff revenues would sum up to more than EUR 5.5 billion during the same time period. 4. ALTERNATIVE SIMULATION RESULTS WITH THE ÖFSE GLOBAL TRADE MODEL In this part, a new assessment of the economic costs and benefits to Canada and the EU based on the ÖFSE Global Trade Model, a structuralist CGE Model, is provided. The dis- tinct difference of our model to standard CGE models is the macroeconomic causality we apply. In our model, output and income are determined by aggregate demand. Another cen- tral feature of the model is that labor market clearance is not required. Therefore, the labor market features unemployment. The pricing power of firms is included as output prices are mark-ups on nominal unit labor costs. Also distributional aspects are considered by way of including a wage bargaining process. However, changes in imports and exports are func- tions of relative prices and demand, in standard fashion. Thus, as in standard CGE models, imports react to changes in trade costs via tariff reductions and NTM alignments (see Raza et al. (2016b) for details). In the following, we briefly discuss the database, model calibration and policy scenarios. In subsequent sections, we present simulation results for four scenarios, discuss the possibil- ity of economic adjustment costs in the transition to the post-liberalization equilibrium, and document various exercises on sensitivity analysis, so as to contextualize results. 4.1. Aggregation, calibration and scenarios The database is disaggregated into eleven countries and regions, and twenty sectors. The eleven regions are five EU countries – Germany, France, Italy, Austria and UK – and two sub-regions, Northern Europe (NE) and Southern & Eastern Europe (SEE). Canada and the US are two further individual countries. Remaining countries are disaggregated into “other OECD” and the rest of the world (ROW). In total, 20 sectors per country/region are included with sectoral breakdowns defined by GTAP (all details included in Table A in the Annex). 13 Source: (01/06/2016)