Bio-based and Applied Economics 7(2): 139-160, 2018 ISSN 2280-6180 (print) © Firenze University Press ISSN 2280-6172 (online) www.fupress.com/bae Full Research Article DOI: 10.13128/bae-7672 The impacts to food consumers of a Transatlantic Trade and Investment Partnership# Yaghoob Jafari1, Wolfgang britz1, JaYson beckman2,* 1 Institute for Food and Resource Economics, University of Bonn, Germany  2 Economic Research Service, USDA, USA Date of submission: 2018 2nd, July; accepted 2019, 29th, January Abstract. Primary agriculture is a textbook example of competitive supply with many producers outputting homogenous products, in contrast to firms in the processed food sector produce heterogeneous products while differing in productivity. Our model of trade reform explicitly accounts for the differences between the markets for primary agriculture and food processing. To demonstrate this point, we use a computable general equilibrium (CGE) model to quantify potential impacts of a trade agreement between the EU and US. Crucially, our heterogeneity-firm setup allows for the allocation of NTMs as ‘fixed costs’, which provides an alternative angle to previous literature that only considered NTM costs in a more conventional framework (e.g., tariff equivalent). Further, the use of this frame- work allows us to provide detailed welfare impacts, providing more information on the impacts to consumers who purchase mainly processed food and little primary agricultural output, a point often unrepresented in previous analysis of NTM reform. Keywords. Trade Policy, Imperfect competition, Heterogeneous firms, Simulations. JEL Codes. F12, F14, F47. 1. Introduction Primary agriculture is a textbook example of competitive supply with many produc- ers outputting homogenous products while firms in the processed food sector produce heterogeneous products. Thus, there are important differences in productivity, size, and exporting behavior among these firms that should be reflected in quantitative analysis. While manufacturing has typically received the firm heterogeneity treatment, the poten- tial Transatlantic Trade and Investment Partnership (TTIP)1 provided an instance of using *Corresponding author: Yaghoob.jafari@ilr.uni-bonn.de # The findings and conclusions in this preliminary publication have not been formally disseminated by the U.S. Department of Agriculture and should not be construed to represent any agency determination or policy. This research was supported in part by the intramural research program of the U.S. Department of Agriculture, Eco- nomic Research Service. 1 We note that the chances for implementation in the near future is slim; however, negotiations have not officially ended (unlike the Trans-Pacific Partnership). The United States Trade Representative still maintains a webpage devoted to the potential agreement. https://ustr.gov/ttip 140 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman an agri-food sector (i.e., food processing) that exhibits firm heterogeneity characteristics. Luckstead and Devadoss (2016) analyze the potential impact of TTIP assuming heteroge- neity of firm involved in food processing. However, that study uses a single sector model that neglects feedback with other sectors, including agriculture as a major upstream link of the processed food sector. The potential TTIP agreement generated a large amount of research, which may not be surprising given that the agreement would have linked the world’s two largest economies. Although the above mentioned paper incorporated a firm heterogeneity setup in their analy- sis of TTIP, most research still considers firms as homogenous (e.g., Arita et al. 2014, 2017; Beckman et. 2015; Beckman and Arita, 2017; Berden et al. 2009; Disdier et al. 2015; Egger et al. 2015; Fontagne et al. 2013; Welfens and Irawan 2014; Beghin et al., 2016). Despite their traditional model approaches, these papers make two points relevant to trade policy anal- ysis. First, tariffs are usually low in developed countries (relative to developing countries), especially for manufacturing and services. However, in developed countries, agriculture is usually more protected relative to other sectors, with higher tariffs, tariff-rate quotas, and non-tariff measures (NTMs).2 Second, for developed countries, NTMs are becoming more of a trade barrier tariffs, with almost all of the papers that compare the two concluding that NTM removal (even partial removal) could generate larger trade gains than those from tar- iff removal. The NTM topic is also relevant for firm heterogeneity, since under the stand- ard Armington assumption, NTMs are typically treated simply as ad-valorem equivalents (AVEs). This differs from the heterogeneous firm layout e.g. used in Akgul et al. (2016) or Luckstead and Devadoss (2016), that treats them (partly) as fixed costs of trade. Our work here builds on the previous TTIP analysis, by starting with the view that reforming agri-food trade might generate the largest relative trade gains. We focus on the processed food sector as it accounts for the largest share of bilateral trade in agri-food between US-EU, and because it can be characterized as exhibiting all the signs of a sector with heterogeneous firms.3 Berden et al. (2009) note that one percent of processed food firms account for 52 percent of total sales. these large firms regularly modify and improve the characteristics of their products to meet the requirements and changing preferences of different consumer groups and to differentiate themselves from competitors.4 Following most TTIP analysis, we employ a CGE model that encompasses all sectors and their inter- actions, but integrate a firm heterogeneity approach for processed food and all other types of manufacturing. This firm heterogeneity specification allows us to provide more infor- mation on impacts to consumers with evidence on the impacts to consumer welfare from a change in the number of new varieties entering the processed food sector, information that is not available in a standard perfect competition setup. Our model also details wel- fare impacts in general, providing information on an aspect of TTIP so far largely ignored. 2 One could examine the average Most Favoured Nation (MFN) tariffs reported to the World Trade Organization to confirm this point. The EU trade-weighted MFN rate for agriculture is 8.7% compared to 2.8% for non-agri- culture. The U.S. rates are 2.3% and 4.0%. The largest gap is likely for Japan: 1.4% and 12.9%. 3 Particularly relevant to TTIP, the EU and US account jointly for one third of global trade in processed food (UN Comtrade, 2015); and they trade (bilaterally) more processed food products than any other partners glob- ally (FAS/USDA, 2014; Olper et al., 2014). 4 In addition, high product differentiation and considerable differences in firms’ size and productivity has led some (e.g., Neff et al., 1996; Francois et al., 2013) to label the processed food sectors as monopolistic competi- tion, which is often used to characterize firms with heterogeneity. 141The impacts to food consumers of a Transatlantic Trade 2. Modeling framework Global CGE models are generally considered well suited for ex-ante appraisal of trade agreements as they consider bilateral trade and trade barriers in a consistent microeco- nomic behavioral framework and account for interlinkages between sectors. Here we use the flexible and modular CGE model by Britz and Van der Mensbrugghe (2017) extended by the heterogeneous firm module of Jafari and Britz (2018a) (see the online appendix for its detailed documentation5). Next, we discuss briefly the general structure of the model. 2.1 Perfect competitive sectors Sectors with perfect competition are depicted as in the standard GTAP model (Hertel, 1997), with cost-minimizing behavior under constant returns to scale (CRS) production technologies along with utility maximizing consumers in competitive markets. Relevant to this work, the perfect competition sectors use an Armington trade setup where a constant elasticity of substitution (CES) function, specific for each agent, i.e. final consumers, gov- ernment, savings and the different production sector, drives competition between domes- tically produced products and imports. A second CES nest, which is not agent specific, depicts the import demand composition from bilateral trade flows. Hence, the Armington setup considers commodities produced in the same region as homogenous, but different from commodities stemming from other regions. For example, all dairy products from the EU are assumed to be of the same quality and fetching the same price. On the sup- ply side, production is defined as the Leontief aggregate of value added and intermediate inputs bundles; the value added composition is based on a CES aggregate of primary fac- tors while the composition of intermediate demand is based on fixed physical input coef- ficients. 2.2 Heterogeneous sectors In each sector there is a continuum of firms that are heterogeneous with regard to productivity, and each firm produces its own distinct variety. While firms are free to enter or exit the market, entrance requires covering fixed costs. Firms learn about their pro- ductivity level once they enter the market, and then choose to stay or exit. Firms with too low of a productivity level will not be able to cover their fixed cost, and therefore exit the industry. For those that survive, only the most productive ones are involved in exports since they can cover the fixed costs of exporting, while less productive firms only serve the domestic market. In this framework, the number of firms operating on the domestic market and on the bilateral trade links depends on the characteristics of the domestic market and bilateral trade costs. Since each firm produces a single distinct variety, the total number of varieties available in any given country depends on the number of firms operating in the domestic market and the number of firms exporting to that country. Accordingly, the total number of varieties available to consumers in a given country is determined endogenously. In this 5 Jafari and Britz (2018) published that online appendix as part of their paper, provided here again for ease. 142 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman context, any policy shock that leads to changes in variable or fixed costs can change the fraction of firms operating on domestic and on trade links, and therefore the number of varieties available to consumers. On the demand side, the composite demand of each agent for each commodity is defined as the Dixit-Stiglitz composite of demand for average firm level varieties around the world6. That index can be interpreted as a standard CES aggregator where the import quantity impact is additionally multiplied by the change in the number of operating firms providing a love of variety effect. Each heterogeneous firm produces one single unique variety and therefore, the num- ber of varieties produced in a regional industry is equal to the number of operating firms. The production structure is shown in Figure 1, where total cost is the sum of variable and fixed costs per firm, the latter consists of fixed costs to enter the enter the industry and fixed cost on each trade link. The variable cost nest uses both primary factors and inter- mediates based on a constant return to scale technology, while fixed cost only relate to primary factors. However, if the overall total cost share of value-added in a sector is small, the fixed cost nest also comprises a share of intermediate composite. This alternative is identified by the intermediate composite in brackets. The value added and intermediate bundles are CES composites of primary factors of production and intermediate inputs, respectively. The total value added (not shown here) is the sum of value added used in both variable and fixed cost nesting. Similarly, the total uses of intermediate commodities and primary factors (not shown here) are the sum of their use in fixed and variable cost nesting. 6 The heterogeneous firm model defines the so-called “average firm” depicting the average productivity of all firms operating on a specific trade link. Figure 1. Production structure in Melitz sectors. Variable cost Commodity 1 … Commodity n Intermediate inputs Value added Factor1 … Factor k Fixed set-up cost Commodity 1 … Commodity n [Intermediate inputs] Value added Factor1 … Factor k Fixed trading cost Commodity 1 … Commodity n [Intermediate inputs] Value added Factor1 … Factor k Total fixed cost Total cost Source: Authors’ illustration. n 143The impacts to food consumers of a Transatlantic Trade Consistent with the monopolistic competition assumption, each firm applies a markup pricing rule, i.e. it collects rent stemming from producing a specific variety, which covers its fixed costs. Marginal production costs are corrected for the average productivity effect of firms operating on each bilateral trade link. The average productivity of firms on each trade link is determined from a Pareto distribution function which encompasses a so-called cut-off productivity level. Only firms with productivity equal to or higher than that specific threshold level for each bilateral trade link will operate on that link, while the remaining firms are forced to exit. The number of operating firms on a link is derived from a zero profit condition where the revenue of the average firm must be equal to its bilateral fixed cost. However, ensuring zero profit for operating firms on each trade link does not ensure zero profits for the industry as a whole, due to the sunk costs associated with the entry of new firms in the industry. Therefore, zero profit at the industry level is assured by a free entry condition in the industry, indicating that the expected profit for firms over their lifetime must be equal to the overall industry fixed set up costs. Trade liberalization filters through this type of model differently than in the stand- ard Armington setup, beginning with the reallocation of resources between firms. For example, a policy that decreases bilateral export cost will encourage some firms that ini- tially did not export (those with low productivity) to start trading. This leads both to an increase in the number of exporters and a decrease in the average productivity of export- ers (since those firms that just entered the export market were less productive to begin with). Due to fixed cost per firm, an increase in the number of exporters implies that the industry as a whole uses more resources. This increases input prices in the domestic mar- ket, leading to some lower productive firms to exit the domestic market. As a result, the average productivity in the domestic market increases. Since some of the least productive firms exit the industry, the productivity for the industry increases and generates a welfare gain (as those firms that now enter the export market are relatively more productive than those leaving the domestic market). On the importing side, similar adjustments in indus- try structure take place while consumer benefits from more varieties being present on the import side. 2.3 Model parameterization and calibration A major advantage of this firm heterogeneity model is that it does not require as much information on industries and consumers as the original Melitz (2003) model. Indeed, only two parameters are needed for each sector: one that describes the productiv- ity distribution of the industry (based on a Pareto distribution) and another that is the elasticity of substitution among domestic and imported varieties. We use the estimate of 3.8 from Bernard et al. (2003) for the elasticity of substitution, and an estimate of 4.6 for the Pareto shape parameter from Balistreri et al. (2011). 2.4 Sectoral and regional aggregation Table A1 in the Appendix provides details on how we treat the sectors in our applica- tion. We generally keep the full sectoral detail of GTAP sectors to prevent bias (Britz et al., 144 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman 2016) but aggregated the processed food sectors in the GTAP data base for two reasons. First, while there is in consensus in the literature that the processed food sector in general should be treated with heterogenous firms, there is no information on if some sectors (e.g., meat or dairy production) should be excluded and treated as homogenous instead. Second, as discussed below in more detail, data on the potential NTM reduction between the EU and the US suitable for our analysis is available only at the aggregated level. To capture the impact of a proposed TTIP agreement on third countries, we aggregate the GTAP data base to 10 regions (European Union, United States, Canada, MERCOSUR, China, ASEAN 10, Mediterranean countries, Other Northern Europe7, low-income coun- tries, Other OECD and Rest of World). Our mapping of regions to the low-income coun- tries aggregate follows the current World Bank classification. 3. Quantifying the policy experiment The model is calibrated based on version 9 of the Global Trade Analysis Project (GTAP) database (Aguiar et al., 2016), which provides a snapshot of world economy in 2011. Figure 2 reports bilateral ad-valorem trade weighted tariffs from the database. It reveals that processed food, beverage and tobacco products, and textile and clothing are the sectors subjected to the highest tariffs. In most cases, the applied rate of the EU is lower than that for the US. NTMs are not explicit in the data base and need to be incorporated before they can be subjected to policy experiments. The AVEs of NTMs that are potentially removable if a deep trade agreement is reached are taken from Egger et al. (2015), the estimated AVE for processed food is 33.83%.8 It is not based on the latest negotiations status of TTIP, but rather more generally reflects the expected change if the two trade partners move to a deep FTA agreement given the empirical evidence from past FTAs. We analyze two scenarios (see Table 1): the first scenario considers completely remov- ing import tariffs for all commodities between the EU and the US, while the second one adds NTM reform. Removing existing tariffs is straightforward as they are part of the data base, whereas the second scenario requires allocating the NTM costs estimated by Egger et al. (2015). CGE models treat the trade cost effects of NTMs as either rent-generating or cost creating. Modeling the rent-generating effect is straightforward using either an “export tax equivalent” – changing export taxes or a “tariff equivalent” approach—chang- ing import taxes, depending on where the rent are assumed to occur. Changes in the cost generating basis of NTMs are modeled by changing the variable portion of trade costs (since there are no fixed costs in an Armington model). However, NTM costs often reflect a ‘fixed cost’ component. For example, the US is able to export beef to the EU, but that beef is produced differently than how most beef sold domestically in the US is produced. To be able to export beef to the EU, US producers must have separate facilities or incur other fixed cost type of costs. As our firm heterogeneity structure is able to account for 7 Other Northern Europe include Switzerland, Norway and Rest of European Free trade Association (EFTA) 8 It should be noted that one would expect the NTMs between the US and the EU to be region specific (i.e., asymmetric). However, Egger et al. (2015) estimated the trade cost equivalents of a deep trade agreement between two regions. Therefore, these estimates should not be interpreted as the current level of NTMs but rath- er as the trade costs that two regions could reduce due to NTM removal when moving to a deep FTA. 145The impacts to food consumers of a Transatlantic Trade fixed cost, we explicitly change variable and fixed costs for EU-US trade links drawing on Jafari and Britz (2018a). One should note that NTMs could also have demand side effects when regulations affect consumer behavior, typically captured by changing either the con- sumer willingness to pay (as in Walmsley and Minor, 2016) or Armington elasticities. Although TTIP might provoke such demand side shifting effects, we leave them out due to missing empirical evidence. Breden et al. (2009) suggested 60% of NTMs in EU-US are cost generating and 40% are rent generating. The later is then allocated by 2/3 to import duties and 1/3 to export taxes following Francoise et al. (2013) and Egger et al. (2015), (see Table 1). The cost por- tion of the NTM is allocated to variable and bilateral fixed costs in equal shares following Jafari and Britz (2018a). 4. Scenario analysis While we presume that costs related to NTMs are already observed in the global SAM, rents related to NTMs probably hide in capital income flows and are clearly so far not allocated bi-laterally. We therefore first run a simulation to include the rent generating effects associated with NTMs currently in place between the US and the EU by introduc- ing respectively increasing bi-lateral import and/or export taxes. That augmented database serves as the benchmark. In the following, we discuss the simulated impacts of both sce- narios on trade, production, and welfare. Then, we turn to the specific outcomes for the food processing sector with a focus on the information given by the firm heterogeneity model. Figure 2. Applied MFN tariff on transatlantic trade. 2 3 2 0 16 6 7 6 2 1 3 2 0 5 1 8 1 1 0 2 4 6 8 10 12 14 16 18 All g oods Grains a nd Crops Livestock and Meat Products Mini ng and E xtraction Processed Food Beverag es and tobacco pr oducts Textiles and Clothing Light Manufacturing Heavy Ma nufacturing EU and US applied tariffs on goods, percent rate US tariffs EU ta riffs Source: Data extracted from version 9 of the Global Trade Analysis Project (GTAP) database (Aguiar et al., 2016). 146 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman 4.1 Effects on trade flows Table 2 shows simulated changes in the volume of aggregate exports. Remov- ing import tariffs (scenario 1) increases EU exports to the US increase by 4.78%, while US exports to the EU increase by 6.8%.9 Adding NTM reform on top of tariff removal boosts bilateral trade further, by 9.7% from the EU to the US, and by 8.5% from the US to the EU. However, with increases of 0.2% (tariff removal only) and 0.5% (NTM reform included), respectively, the changes in global EU exports are minor; while total US exports expand more significantly (by 1.4% and 2.6%). These findings are comparable with Fran- cois et al. (2013). Some regions including China, ASEAN 10, “low-income countries”, and “Other Northern Europe” have marginally increases in their exports to either the EU or US, depending on the scenario. In the first scenario, Canada has a decrease in their bilat- eral exports to both regions, but in the second scenario, Canada has an increase in their exports to the EU. In summary of these changes in regional trade flows, overall world trade increases marginally by 0.3% and 0.4%, respectively. Table 3 focuses on export flows for the processed food sector. The higher tariff pro- tection in that sector leads to larger changes compared to the results reported above: EU exports to the US of processed food increase by 39% while US exports to the EU increase by 121% (for tariff removal). These findings are consistent with the partial-equilibrium model results of Luckstead and Devadoss (2016), but the magnitude of the impacts found here is different due to the use of different elasticities and feedback effects in our CGE modelling. As the AVE estimates of the expected changes in existing NTMs between the 9 Bilateral changes are not presented here, but are available upon request from the authors. Table 1. Scenario layout. Tariffs shocks AVEs shocks Total AVEs reduction divided into the last three columns Import tax Export tax Bilateral fixed and variable trade cost (1) (2) (2)*0.4*2/3 (2)*0.4*1/3 (2)*0.6 Scenario 1 -100% reduction for all economic sectors - - - - Scenario 2 -100% reduction for all economic sectors -33.83% -9.0% -4.50% -20.3% Modeled as Reduction in bilateral import tariff Reduction in import tariffs representing rents in importer country Reduction in export taxes representing rents in exporter country Converted to an equivalent reduction in bilateral fixed and variable trade cost Source: authors. 147The impacts to food consumers of a Transatlantic Trade EU and US are quite high and exceed existing tariff levels, bilateral trade volumes increase considerably for processed food in the second scenario. EU exports to the US almost quadruple, while US exports to the EU multiply by more than seven. This leads to changes in total exports of processed food for the EU by almost 9% and by 63% for the US. The trade diversion effects of that second scenario in the processed food sector is accordingly sizeable: most EU trading partners lose about 4% of their exports while exports to the US from the non-EU countries decreases by around 10%. Trade impacts for primary agriculture are minor (see Table A2 and A3) which reflects low tariffs (see Figure 2) and low exports values. EU exports of primary agricultural prod- ucts to the US amount to about 83 million, vice versa it is 6 million. Our analysis also shows that the impact on average manufacturing trade between two regions is small (see Table 4.4) due to low tariffs between the regions. 4.2 Effects on domestic output quantities Table 4 presents information on production changes across all sectors. For processed food, the EU faces a decrease in both scenarios, while the US increases its production. However, the increase in US production is small, as the 63% increase in exports is mostly offset by an increase in imports (46 %) (See Table 3). Opposite and stronger effects are simulated for beverages and tobacco, with a 5% increase in EU production and a 16% Table 2. Change in aggregate exports by region [% change]. Regions Scenario 1 Scenario 2 EU US Total EU US Total World 0.2 1.1 0.3 0.4 1.8 0.4 EU -0.21 4.8 0.2 -0.3 9.7 0.5 Other Northern Europe 0.1 0.2 0.2 0.0 -0.4 -0.1 US 6.8 1.4 8.5 2.6 Canada -0.1 -0.3 -0.2 1.3 -0.1 0.2 Mercosur -0.2 0.1 0.0 -0.1 -0.5 -0.1 China 0.0 0.3 0.2 0.5 -0.5 0.1 ASEAN 10 -0.4 0.5 0.7 0.2 -0.7 0.2 Other OECD -0.2 0.0 0.0 0.5 -0.6 0.0 EU Mediterranean Partners -0.3 0.3 -0.1 0.1 -0.5 0.0 Low Income -0.3 0.3 -0.1 0.1 -0.1 0.0 Rest of World -0.2 0.1 0.0 0.1 -0.5 0.0 Notes: exporters in rows, importers in columns. Source: model results. 1 The reader should note that the numbers presented in the column “EU” showing EU to EU exports is due to an aggregation effect. Sales to the domestic market of a nation are not reported as exports in the SAM. However, if we aggregate individual EU countries, the former bi-lateral trade links between two EU nations occur now inside one aggregate and become the diagonal trade flow in this column. The domestic sales of the EU aggregate are defined from adding up the domestic sales of individual EU countries. 148 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman decrease in the US. This happens for two reasons: 1) the EU has larger base exports of beverages and tobacco relative to the US; 2) the US has relatively higher tariffs on bever- ages and tobacco compared to processed food. Other sectors of the economy show only marginal changes. An exception is the output of “Textiles and Clothing”, which has a 2.5% increase in the EU in the first scenario. This gain disappears in the second scenario as resources flow to beverages and tobacco in order to meet the large increase in production. Overall, the domestic output of processed food sectors in the EU is simulated to increase by 1.4% in the second scenario, while US output drops by 2.9%. This result is different from that found in other TTIP studies. Those studies generally conclude that the US has large production gains at the expense of the EU. 4.3 Effects on welfare Welfare impacts are measured based on the equivalent variation (EV) criterion, i.e., the amount of money to be added to the regional household’s benchmark income at benchmark prices to reach the same utility as under simulated income and prices. There are global welfare gains of 5.6 billion USD when tariffs are removed (see Table 5), of which 2.8 billion USD accrue to the EU and 5 billion USD to the US (the results are com- parable with Francois et al., 2013); the remaining countries, with the exception of China, have losses below 1 billion USD. Both changes in the intensive and extensive margin of trade are important in determining the welfare changes in other countries: Following a reduction in trade barriers between the US and EU, the intensive margin of trade between the two regions increases, diverting trade with other countries and causing welfare to decrease. However, a reduction in trade barriers between the EU and US helps increase the average productivity of firms operating on the domestic market and/or operating on Table 3. Export volumes by region for “processed food” [% change]. Regions Scenario 1 Scenario 2 EU US Total EU US Total World 1.5 5.3 1.1 7.2 46.2 7.4 EU 0.2 39.4 1.2 -1.7 394.2 9.3 Other Northern Europe -0.7 -0.1 -0.4 -5.4 -10.8 -4.4 US 120.9 9.4 748.5 63.4 Canada -1.0 -0.4 -0.4 -5.7 -11.3 -8.6 Mercosur -0.7 0.0 -0.2 -3.9 -9.1 -1.6 China -0.7 0.0 -0.1 -3.9 -9.3 -2.4 ASEAN 10 -1.0 -0.3 -0.3 -3.7 -8.9 -1.7 Other OECD -0.7 -0.1 -0.1 -3.8 -9.3 -2.7 EU Mediterranean Partners -0.6 0.0 -0.3 -4.5 -9.7 -2.5 Low Income -0.6 0.0 -0.3 -4.5 -9.5 -2.5 Rest of World -0.7 0.0 -0.2 -4.1 -9.4 -2.1 Source: model results. 149The impacts to food consumers of a Transatlantic Trade trade links other than EU-US trade link. This results in an increase in the intensive mar- gin of trade (i.e., increase in varieties) in other countries, which is welfare increasing. The total welfare impact on third countries is therefore determined based on the total volume of trade, i.e., the sum of changes in intensive and extensive margins of trade. All regions are better off compared to the first scenario if NTMs are also reduced, sev- eral regions besides the EU and US now experience welfare increases, which results in a global welfare gain of 22.4 billion USD. The removal of NTMs increases average domes- tic productivity, simulating the extensive trade margin, and improving welfare compared to the first scenario. Still, welfare losses occur in Canada, Mercosur, ASEAN 10, and oth- er OECD countries. The EU has the largest additional welfare gains, increasing from 2.8 billion USD to 13.8 billion USD under the second scenario. The US has an additional 5 billion USD added in the second scenario to reach a total of 7.8 billion USD. The welfare improvements in the second scenario match findings by Balistreri et al. (2011) who reports that NTM reduction in the Melitz (2003) framework increases welfare considerably. Further, our welfare decomposition analysis reveals that the largest portion of welfare gains are associated with the scale effect (associated with the increase in returns to scale), the productivity effect (expansion in market shares of efficient firms), and variety effects (i.e., increases in the number of varieties face by consumers). While term of trade and allocative efficiency contribution is small, the fixed cost effect (due to the increase in firms fixed cost payments) reduces welfare (Table 6). 4.4 Firm-level impact of policy shocks in processing food sectors Table 7 shows the change to the average firm (as shown in rows) associated with the production and sale of processed food in the EU for different bilateral trade markets. The Table 4. Industrial output by sector [% change]. Sectors EU US Scenario 1 Scenario 2 Scenario 1 Scenario 2 Total 0.00 -0.01 0.02 0.00 Processed food -0.11 -0.05 0.30 0.46 Beverages and Tobacco 0.13 5.67 -0.23 -16.00 Grains and Crops -0.13 0.29 0.23 -0.24 Livestock -0.08 0.03 0.22 0.23 Mining and Extraction -0.01 -0.05 -0.05 0.07 Textiles and Clothing 2.52 -0.01 0.28 1.11 Light Manufacturing -0.09 -0.24 0.51 0.15 Heavy Manufacturing -0.13 -0.40 -0.25 0.30 Utilities and Construction 0.00 0.00 0.05 -0.01 Transport and Communication 0.02 0.07 0.01 0.02 Other Services -0.01 0.01 -0.01 0.02 Source: model results. 150 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman first column refers to the domestic market, the second column denotes intra-EU trade, the third and fourth columns show EU trade with the US and other regions not included in the transatlantic trade block (hereafter referred as nonTTIP). The last column relates to overall industry performance. Table 5. Changes in welfare [Billion USD]. Regions Scenario 1 Scenario 2 World 5.6 22.4 EU 2.8 13.8 Other Northern Europe -0.2 0.1 US 5.0 7.8 Canada -0.1 -0.1 Mercosur -0.1 -0.1 China 0.0 0.6 ASEAN 10 -0.2 -0.1 Other OECD -0.7 -0.2 EU Mediterranean Partners -0.2 0.2 Low Income -0.1 0.1 Rest of World -0.7 0.3 Source: model results. Table 6. Welfare decomposition analysis. Scenario 1 Scenario 2 EU 2.8 13.8 Allocative efficiency 0.0 0.5 Term of trade effect 0.1 1.1 Variety effect 1.3 5.2 Scale effect 1.9 9.7 Productivity effects 0.9 4.2 Fixed cost effects -1.3 -6.9 Other effects -0.1 0.0 US 5.6 22.4 Allocative efficiency 0.3 0.9 Term of trade effect -0.1 1.7 Variety effect 1.8 4.7 Scale effect 3.6 11.7 Productivity effects 2.6 8.5 Fixed cost effects -2.3 -5 Other effects -0.2 -0.1 Source: Authors’ calculations based on model results. 151The impacts to food consumers of a Transatlantic Trade Table 7. Average firm results for EU domestic sales and exports of processed food [% change]. Scenario 1 Scenario 2 Domestic sales EU US nonTTIP Total sale Domestic sales EU US nonTTIP Total sale Firm price -0.2 -0.1 9.5 0.0 0.8 -0.9 -0.9 17.6 0.0 1.3 Number of operating firms -0.6 0.1 52.7 0.0 4.3 -2.9 -2.7 666.5 0.9 55.8 Avg. output per firm 0.2 0.1 -8.7 0.1 0.1 1.0 1.0 -35.5 0.1 0.8 Avg. productivity per firm 0.1 0.0 -8.8 0.0 -0.9 0.7 0.7 -35.7 0.0 -2.9 Industry Fix costs 0.0 0.0 0.0 0.0 0.0 0.1 0.1 -24.2 0.0 -0.1 Fix costs per unit 0.3 -0.2 -28.3 0.0 -2.1 2.0 1.9 -84.7 0.0 -7.3 Industry Variable costs -0.6 0.1 52.7 0.0 -0.2 -3.0 -2.6 481.2 1.2 -0.4 Variable costs per unit -0.3 -0.1 9.5 0.0 0.8 -1.1 -0.9 17.6 0.0 1.2 Total output sold -0.4 0.2 39.4 0.1 0.0 -1.9 -1.7 394.2 1.3 0.7 Source: Based on model results. The changes in the EU-US trade link for the tariff removal scenario shows a typical reaction of the firm heterogeneity model: tariff removal reduces the average import price in the US, allowing less productive EU firms to operate on that trade link. This increas- es the number of firms and varieties exported to the US (52%), providing benefits to US consumers. Per unit fix costs drops by 28%; however, lower average productivity increases the variable costs per unit by about 9.5%. There is an increase in total output sold to the US of 39%, but increasing the number of operating firms decreases the average productiv- ity of the firms operating on that trade link (-8.8%).10 Thus, the average size of these firms also drops – average output per firm decreases by about the same percentage. The aver- age firm exporting to the US after these changes is less productive and smaller. Together, these changes constitute a new equilibrium with zero profits for the firms operating on that trade link, while monopolistic prices charged are equal to the willingness to pay for the specific quality delivered on that trade link given the number of varieties available. The impacts of the second scenario on EU-US bilateral trade are more pronounced: besides tariff removal, we also shock variable and fix costs related to NTMs for EU-US bilateral trade. This amplifies the effect compared to the first scenario, as now all firms face a higher willingness to pay in bilateral trade, and experience cost savings before sup- ply and demand adjust. This allows far less efficient firms to operate in bilateral trade: the number of the EU firms exporting to the US increases by 666%11 while average produc- tivity (35%) and firm size (-35%) on the trade link drop. Average per unit variable costs increase by 18%, which translates into changes in the average firm price, while total out- put for EU-US bilateral trade almost quadruples. The fix cost of the industry operating on 10 Note that in Table 7 and subsequent tables, even though the number of operating firms increases, the total out- put change is small because each firms now produces less output. This is equivalent to saying that large increases in the extensive margin are compensated by a reduction in the intensive margin of trade. 11 Only a small share of firms operate on the link before trade liberalization which are the firms with the highest productivity. Given the shape of the productivity distribution, a significant decrease in bilateral trade cost leads to a nover-proportional increase in that share and increase at the same also the number of traded varieties on the link. 152 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman that link decreases by 24%, reflecting our assumption of reduced trade costs (see Table  1). However, the original reduction is partly offset by the loss in average productivity, and at the same time distributed to a much higher output quantity. The combined impact on per unit fix costs on that link drops by around -85%. The finding is in line with the literature emphasizing the importance of the extensive margin of trade (e.g., Hummel and Klenow, 2005; Chaney, 2008; among others). No significant changes occur on the EU-nonTTIP link, such that overall changes in trade reflect only the discussed EU-US bilateral changes. The expansion in exports combined with an on average less productive firm that trades, increases the overall input demand in the economy. This in turn bids up factor and other intermediate prices. As a first order impact, production costs increase and profits on other trade links decline, which induces some of the less productive firms to exit the EU domestic market. The number of operating firms in the domestic market decreases by 0.6% and 2.9% in the first and second scenario. As firms with lower productivity exit, factors are reallocated towards higher-productive and larger firms, thus the average pro- ductivity of firms operating in the domestic market rises by 0.1% and 0.7%. This leads to a decline in variable per unit costs of 0.3% and 1.1%, and an increase in average output per firm of 0.2% and 1%. However, the increase in average firm output does not compensate for the decrease in the number of firms operating in the domestic market. Consequently, domestic sales decline by 0.4% and 1.9%. This, along with lower firm prices of -0.2% and -0.9%, reflects the increased competition with US imports. The impact on export flows of processed food from the US to the EU is presented in Table 8. Note first the impact on US-EU trade: following the reduction in border protec- tion and trade cost, less productive firms find it profitable to enter the trade market. Thus, the number of operating firm on the US-EU link increases by a factor of 1.7 in the first scenario and by 1412 in the second scenario. This lowers average productivity on that link, such that there are increases in the average firm price and output. Still, US exports to the 12 Please see footnote 12. Table 8. Average firm results for US domestic sales and exports of processed food [% change]. Scenario 1 Scenario 2 Domestic sales EU nonTTIP Total sale Domestic sales EU nonTTIP Total sale Firm price -0.1 24.6 0.0 2.3 -2.0 35.6 0.0 3.1 Number of operating firms -0.6 174.9 -0.9 15.1 -5.0 1421.0 4.0 132.0 Avg. output per firm 0.3 -19.6 0.3 0.2 1.9 -44.2 0.0 1.6 Avg. productivity per firm 0.2 -19.7 0.0 -1.9 1.4 -44.5 0.0 -4.1 Industry Fix costs 0.1 0.1 0.0 0.1 -0.1 -24.4 0.0 -0.2 Fix costs per unit 0.4 -54.7 0.0 -4.6 3.1 -91.1 0.0 -8.0 Industry Variable costs -0.4 175.2 -0.4 0.4 -5.5 1050.5 3.6 -0.4 Variable costs per unit -0.1 24.6 0.0 2.2 -2.5 35.6 0.0 2.9 Total output sold -0.3 120.9 -0.3 0.4 -3.1 748.5 3.7 1.3 Source: Based on model results 153The impacts to food consumers of a Transatlantic Trade EU increase considerably (by a factor of 1.2 and 7.5), which reflects tariff removal plus an increased willingness to pay due to a higher number of varieties. Export expansion ulti- mately negatively affects the output sold in the domestic market by 0.3% in the first sce- nario and 3% in the second. Accordingly, total US processed food sales increases by only 0.4% (in the first scenario) and 1.3% (in the second). 5. Sensitivity analysis The policy shock, model structure and parameterization jointly determine the model results. We check their robustness with regard to welfare and the volume of exports in the processed food sector. Given the uncertainties on the future of TTIP negotiations, we first perform a sensitivity analysis with regard to tariff and NTM reduction. To do so, we impose a 50% tariff shock (similar to Francois et al., 2013) instead of the 100% removal in the benchmark. This essentially takes into account agricultural products that could be exempted from tariff removal. Our results (not shown here) indicate negligible impacts on trade in processed food and welfare and a small impact on overall primary agriculture trade. We also perform a sensitivity analysis for the NTM reduction scenario, but allowing for only half of the reduction in NTM costs. Figure 3 shows that the simulated changes in trade in processed food are 75-95% and welfare gains are 27-30% lower compared to the earlier results. Next, we compare changes in welfare and trade in processed food in the tariff/NTM removal between model setups, i.e. proceed food and all manufacturing sectors have the firm heterogeneity setup (MEL), and the more conventional structure where all the sectors follow the standard Armington specification (ARM). Welfare in ARM scenario is about 40-50% and trade effects are 10-30% lower compared to the firm heterogeneity configura- Figure 3. Change in welfare [Equivalent variant per capita in constant USD] and export volumes [Mil- lion constant USD] under lower reduction of NTMs. 0 5 10 15 20 25 World EU US U SD p er c ap ita 33.8% r eduction in NTMs 16.9% r eduction in NTMs 0 100 200 300 400 500 600 700 800 900 World EU US M ill io n U SD 33.8% r eduction in NTMs 16.9% r eduction in NTMs Source: Simulation results. 154 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman tion (Figure 4). Comparable relative differences in welfare and trade flows are reported by Hosoe (2017), and Jafari and Britz (2018b) for a simulation of Brexit in a CGE model. Lastly, an additional sensitivity analysis shows that trade expansion and welfare gains are higher under a higher shape parameter, i.e. if the distribution of firms’ productivity becomes steeper. In Figure 5, we compare the results when the shape parameter is one- third higher than the benchmark value. The gain in welfare is 30-80% and trade in pro- cessed food is 20-40 % higher than under the default parameters across different regions. The results are comparable with Zhai (2008), who simulated a 50% reduction in manufac- Figure 4. Change in welfare [Equivalent variant per capita in constant USD] and export volumes [Mil- lion constant USD] under the under Melitz and Armington specification. 0 5 10 15 20 25 World EU US U SD p er c ap ia MEL ARM 0 200 400 600 800 1000 World EU US U SD m ill io n MEL ARM Source: Simulation results. Figure 5. Change in welfare [Equivalent variant per capita in constant USD] and export volumes [Mil- lion constant USD] under different shape parameters of the Pareto distribution of firm productivity. 0 5 10 15 20 25 30 World EU US U SD p er c ap ita Benchmark shape para meter (4. 8) Higher shape para meter (6.4) 0 100 200 300 400 500 600 700 800 900 1000 World EU US U SD m ill io n Benchmark shape para meter (4. 8) Higher shape para meter (6.4) Source: Simulation results. 155The impacts to food consumers of a Transatlantic Trade turing tariffs across the world. We also test the implication of increasing the benchmark Armington elasticities by one-third but keeping the original shape parameter. Our results (not shown here) reveals that that under this assumption, exports are about 10-12% lower across regions, with only a modest impact 4-5% increase in welfare. 6. Conclusion This study employs a CGE model with a firm heterogeneity setup for processed food and manufacturing to simulate impacts of a potential TTIP agreement on the EU, the US and other countries. This setup allows us to trace the impacts on the intensive and extensive margin of trade as well as on firm productivity. In addition, in accounting for firm hetero- geneity by allowing fixed costs to vary, we can more flexibly allocate NTM compared to the more conventional Armington set-up. We simulate the impacts of (i) removing all bilat- eral tariffs currently in place between the EU and the US; and (ii), an additional removal of NTMs in food processing sectors. Dismantling bilateral import tariffs leads to bilateral trade impacts that are below +10%, and limited welfare and trade diversion effects. As empirical estimates in the previous literature of the welfare impacts of NTMs suggest that these form considerable barriers, and the results of our second scenario are consistent with those of the earlier studies. In particular, EU welfare increases from 2 billion USD under the first sce- nario to 13.8 billion USD under the second. The larger increase in exports for food process- ing stems almost entirely from more firms exporting, which underlines the importance of the Melitz model in the analysis. However, increased exports are offset by lower domestic sales in both regions, such that overall industry output changes little. A sensitivity analysis on the core parameters used in the model shows the robustness of the overall results. Our results differ from previous analysis of TTIP, which suggested that the US would see larger increases than the EU in agri-food production and exports. While this is still somewhat the case for primary agriculture, our results indicate that the EU could see larg- er production gains for processed food with the U.S. experiencing a decrease in output. Our results also indicate that tariff removal alone could benefit both regions once produc- tivity gains are considered—something that cannot be shown with the typical Armington model setup. In the end, consumers could benefit most from a TTIP agreement as prices are likely to fall and the diversity of products available increase. Our study could only draw on rather aggregate estimates of the costs caused by NTMs in the food processing sector. Thus, as the sector is highly heterogeneous, future work could try to provide more disaggregated estimates of costs related to NTMs and their composition (rents in importer and exporter country, variable or fixed cost of trade, demand shifting etc.). That would clearly not only improve the analysis of a potential TTIP agreement, but more generally economic impact assessment of FTAs and multilat- eral trade liberalization. 7. Acknowledgments The authors would like to thank Thomas Heckelei, Peter Witzke, and Valentina Rai- mondi (the editor of the journal), and anonymous referees for their insightful comments that greatly improved this paper. 156 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman 8. References Aguiar, A., Narayanan, B. and McDougall, R. (2016). An Overview of the GTAP 9 Data Base. Journal of Global Economic Analysis 1 (1): 181-208. Akgul, Z., Villoria, N.B. and Hertel, T.W. (2016). GTAP-HET: Introducing Firm Heteroge- neity into the GTAP Model. Journal of Global Economic Analysis 1 (1): 111-180. Arita, S., Beckman, J., Kuberka, L. and Melton, A. (2014). Sanitary and Phytosanitary Measures and Tariff Rate Quotas for U.S. Meat Exports to the European Union. http://www.ers.usda.gov/publications/pub-details/?pubid=37588. Accessed 15 November 2017. Arita, S., Beckman, J. and Mitchell, L. (2017). Reducing Transatlantic Barriers on U.S.-EU Agri-Food Trade: What Are The Possible Gains?. Food Policy 68: 237-247. Balistreri, E.J., Hillberry, R.H. and Rutherford, T.F. (2011). Structural Estimation and Solu- tion of International Trade Models with Heterogeneous firms. Journal of Internation- al Economics 83(2): 95–108. Beckman, J. and Arita, S. (2017). Modeling the Interplay Between Sanitary and Phytosani- tary Measures and Tariff-Rate Quotas under Partial Trade Liberalization. American Journal of Agricultural Economics 99(4): 1078-1095. Beckman, J., Arita, S. and Mitchell, L. (2015).The Transatlantic Trade and Investment Partnership and Agriculture: A Quantitative Analysis. Agricultural & Applied Eco- nomics Association Conference Paper. Beghin, J.C., Bureau, J.C. and Gohin, A. (2016). The Impact of an EU-US Transatlantic Trade and Investment Partnership Agreement on Biofuel and Feedstock Markets. Journal of Agricultural Economics 68: 321-344. Berden, K.G., Francois, J., Thelle, M., Wymenga, P. and Tamminen, S. (2009). Non-Tariff Measures In EU-US Trade And Investment – An Economic Analysis. http://trade. ec.europa.eu/doclib/docs/2009/december/tradoc_145613.pdf. Accessed 15 Novem- ber 2018. Britz, W. and Van der Mensbrugghe, D. (2018). The Standard GTAP Model in GAMS, Version 7. Journal of Global Economic Analysis 3(1): 1-83. Britz, W. (2017). CGEBox—A Fexible and Modular Toolkit for CGE Modelling with a GUI. University of Bonn, Bonn. http://www.ilr.uni-bonn.de/em/rsrch/cgebox/cge- box_GUI.pdf. Accessed 25 September 2017. Britz W. and Van der Mensbrugghe, D. (2016) Reducing Unwanted Consequences of Aggregation in Large-Scale Economic Models - A Systematic Empirical Evaluation with the GTAP Model. Economic Modelling 59: 462-473. Britz W. and Van der Mensbrugghe, D. (2017). A Flexible, Modular and Extendable Framework for CGE Analysis in GAMS. 20th GTAP Conference. 59: 462–473. Chaney, T. (2008). Distorted Gravity: The Intensive and Extensive Margins of Internation- al Trade. American Economic Review 98 (4): 1707-1721. Disdier, A.C., Emlinger, C. and Fouré, J. (2016). Interdependencies between Atlantic and Pacific Agreements: Evidence from Agri-Food Sectors. Economic Modelling 55: 241- 253. Egger, P., Francois, J. Manchin M. and Nelson, D. (2015). Non-Tariff Barriers, Integration and the Transatlantic Economy. Economic Policy 30 (83): 539-584. 157The impacts to food consumers of a Transatlantic Trade FAS/USDA. (2014). Agricultural Exports to the European Union: Opportunities and Chal- lenges. http://www.fas.usda.gov/data/agricultural-exports-european-union-opportunities-and- challenges. Accessed 15 March 2018. Fontagné, L., Gourdon, J. and Jean, S. (2013).Transatlantic Trade: Whither Partnership, Which Economic Consequences?. http://www.cepii.fr/PDF_PUB/pb/2013/pb2013- 01.pdf. Accessed 15 March 2018. Francois, J., Manchin, M., Norberg, H., Pindyuk O. and Tomberger, P. (2013). Reducing Transatlantic Barriers to Trade and Investment – An Economic Analysis. http:// trade.ec.europa.eu/doclib/docs/2013/march/tradoc_150737.pdf. Accessed 15 March 2018. Hertel, T.W. (1997). Global Trade Analysis: Modelling and Applications. Cambridge Uni- versity Press. Hertel, T.W., Walmsley, T.L. and Itakura, K. (2001). Dynamic Effects of the “New Age” Free Trade Agreement between Japan and Singapore. Journal of Economic Integra- tion 16: 446-484. Hosoe, N. (2017). Impact of Border Barriers, Returning Migrants, And Trade Diversion In Brexit: Firm Exit And Loss Of Variety. Economic Modelling 69: 193-204. Hummels, D. and Klenow, P.J. (2005).The Variety and Quality of a Nation’s Exports. Amer- ican Economic Review 95(3): 704-723. Jafari Y. and Britz, W. (2018a). Modelling Heterogeneous Firms and Non-Tariff Measures in Free Trade Agreements Using Computable General Equilibrium. Economic Mod- elling 73: 279-294. Jafari Y. and Britz, W. (2018b). Brexit: An Economy-Wide Impact Assessment on Trade, Immigration, and Foreign Direct Investment. Empirica https://doi.org/10.1007/ s10663-018-9418-6. Luckstead, J. and Devadoss, S. (2016). Impacts of the Transatlantic Trade and Investment Partnership on Processed Food Trade under Monopolistic Competition and Firm Heterogeneity. American Journal of Agricultural Economics 98(5): 1389-1402. Melitz, M.J. (2003). The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity. Econometrica 71(6): 1695-1725. Neff, S., Harris, M., Malanoski, M. and Ruppei, F. (1996). U.S. Trade In Processed Foods. In globalization of the processed foods market, ed. D.R. Henderson, C.R. Handy, and S.A. Neff, 25–65 (1996). USDA, Economic Research Service, Washington, D.C. Olper, A., Curzi, D. and Pacca, L. (2014). Trade, Import Competition and Productivity Growth in the Food Industry. Food Policy 49(1): 71-83. Van der Mensbrugghe, D. (2018). The Standard GTAP Model in GAMS, Version 7. Jour- nal of Global Economic Analysis 3(1): 1-83. Welfens, P. and Irawan, T. (2014). Transatlantic Trade and Investment Partnership: Sec- toral and Macroeconomic Perspectives for Germany, the EU and the US. Interna- tional Economics and Economic Policy 11: 293-328. Walmsley, T.L. and Minor, P. (2016). Willingness to Pay in CGE Models, ImpactECON Working Paper No. 04, ImpactECON: Boulder, CO, USA. Zhai, F. (2008). Armington Meets Melitz: Introducing Firm Heterogeneity in a Global CGE Model of Trade. Journal of Economic Integration 23(3)3: 575-604. 158 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman Appendix. Supplemental tables Table A1. Sectoral correspondence of GTAP 9 sector to new sectors. Number Code Description Pre model Aggregation Post model aggregation Market Structure 1 PDR Paddy rice Paddy rice Grains and Crops PC 2 WHT Wheat Wheat Grains and Crops PC 3 GRO Cereal grains nec Cereal grains nec Grains and Crops PC 4 V_F Vegetables, fruit, nuts Vegetables, fruit, nuts Grains and Crops PC 5 OSD Oil seeds Oil seeds Grains and Crops PC 6 C_B Sugar cane, sugar beet Sugar cane, sugar beet Grains and Crops PC 7 PFB Plant-based fibers Plant-based fibers Grains and Crops PC 8 OCR Crops nec Crops nec Grains and Crops PC 9 CTL Bovine cattle, sheep and goats, horses Bovine cattle, sheep and goats, horses Livestock PC 10 OAP Animal products nec Animal products nec Livestock PC 11 RMK Raw milk Raw milk Livestock PC 12 WOL Wool, silk-worm cocoons Wool, silk-worm cocoons Livestock PC 13 FRS Forestry Forestry Mining and Extraction PC 14 FSH Fishing Fishing Mining and Extraction PC 15 COA Coal Coal Mining and Extraction PC 16 OIL Oil Oil Mining and Extraction PC 17 GAS Gas Gas Mining and Extraction PC 18 OMN Minerals nec Minerals nec Mining and Extraction PC 19 CMT Bovine meat products Processed food Processed food FH 20 OMT Meat products nec 21 VOL Vegetable oils and fats 22 MIL Dairy products 23 PCR Processed rice 24 SGR Sugar 25 OFD Food products nec 26 B_T Beverages and tobacco products Beverages and tobacco products Beverages and tobacco products FH 27 TEX Textiles Textiles Textile and clothing FH 28 WAP Wearing apparel Wearing apparel Textile and clothing FH 29 LEA Leather products Leather products Light Manufacturing FH 30 LUM Wood products Wood products Light Manufacturing FH 31 PPP Paper products, publishingPaper products, publishing Light Manufacturing FH 32 P_C Petroleum, coal products Petroleum, coal products Heavy Manufacturing FH 159The impacts to food consumers of a Transatlantic Trade Number Code Description Pre model Aggregation Post model aggregation Market Structure 33 CRP Chemical, rubber, plastic products Chemical, rubber, plastic products Heavy Manufacturing FH 34 NMM Mineral products nec Mineral products nec Heavy Manufacturing FH 35 I_S Ferrous metals Ferrous metals Heavy Manufacturing FH 36 NFM Metals nec Metals nec Heavy Manufacturing FH 37 FMP Metal products Metal products Light Manufacturing FH 38 MVH Motor vehicles and parts Motor vehicles and parts Light Manufacturing FH 39 OTN Transport equipment nec Transport equipment nec Light Manufacturing FH 40 ELE Electronic equipment Electronic equipment Heavy Manufacturing FH 41 OME Machinery and equipment nec Machinery and equipment nec Heavy Manufacturing FH 42 OMF Manufactures nec Manufactures nec Light Manufacturing FH 43 ELY Electricity Electricity Utilities and Construction PC 44 GDT Gas manufacture, distribution Gas manufacture, distribution Utilities and Construction PC 45 WTR Water Water Utilities and Construction PC 46 CNS Construction Construction Utilities and Construction PC 47 TRD Trade Trade Transport and Communication PC 48 OTP Transport nec Transport nec Transport and Communication PC 49 WTP Water transport Water transport Transport and Communication PC 50 ATP Air transport Air transport Transport and Communication PC 51 CMN Communication Communication Transport and Communication PC 52 OFI Financial services nec Financial services nec Other Services PC 53 ISR Insurance Insurance Other Services PC 54 OBS Business services nec Business services nec Other Services PC 55 ROS Recreational and other services Recreational and other services Other Services PC 56 OSG Public Administration, Defense, Education, Health Public Administration, Defense, Education, Health Other Services PC 57 DWE Dwellings Dwellings Other Services PC Notes: FH: Firm heterogeneity, PC: Perfect Competition (Armington). 160 Yaghoob Jafari, Wolfgang Britz, Jayson Beckman Table A2. Export volume by region for “crop products” [% change]. Exporters Partners EU Other Northern Europe US Canada Mercosur China ASEAN 10 Other OECD EU Mediterranean Partners Low Income Rest of World First scenario World 0.3 0.0 1.4 -0.2 -0.2 -0.3 -0.2 -0.2 -0.3 -0.2 -0.3 EU -0.3 0.0 16.4 0.5 -0.2 0.0 0.0 0.4 -0.1 -0.1 0.0 US 15.1 -1.1 -0.7 -1.3 -1.2 -1.1 -0.7 -1.2 -1.2 -1.2 Second Scenario World 0.5 -1.8 2.8 -1.7 -0.6 -0.5 -1.2 -0.8 -0.4 -0.3 -0.5 EU -0.1 -1.7 17.4 -0.1 -0.7 0 -0.8 0.2 -0.1 -0.2 -0.1 US 14.1 -4 0 -2.8 -2.9 -2.5 -3.2 -2.1 -2.4 -2.5 -2.5 Table A3. Export volume by region for “livestock products” [% Change]. Exporters Partners EU Other Northern Europe US Canada Mercosur China ASEAN 10 Other OECD EU Mediterranean Partners Low Income Rest of World First scenario World 0.2 -0.2 1.3 0.0 -0.1 -0.2 -0.1 0.0 -0.2 -0.1 -0.1 EU 0.0 -0.2 9.9 0.2 -0.2 -0.1 -0.1 0.2 -0.2 -0.1 -0.1 US 11.5 -0.5 -0.1 -0.6 -0.4 -0.4 -0.2 -0.5 -0.4 -0.4 Second Scenario World -0.5 -0.9 -2.2 4.9 -0.1 0 -0.2 1.2 0.2 -0.1 0 EU -0.6 -0.8 9.4 3.4 0 0.3 0 0.9 0.2 0.3 0.2 US 13.5 1.5 0 5.8 2.1 2.6 2.2 3 2.5 2.5 2.4 Table A4. Export volume by region for overall manufacturing sectors [% Change]. Exporters Partners EU Other Northern Europe US Canada Mercosur China ASEAN 10 Other OECD EU Mediterranean Partners Low Income Rest of World First scenario World 0.3 -0.1 1.3 0.0 -0.2 -0.1 -0.1 -0.1 -0.1 0.0 0.0 EU -0.2 0.0 3.9 0.3 -0.1 0.0 0.1 0.1 0.1 0.1 0.1 US 5.2 -0.4 -0.2 -0.5 -0.5 -0.4 -0.4 -0.4 -0.3 -0.4 Second Scenario World 0.4 -0.1 1.4 -0.4 -0.4 -0.1 0.0 -0.1 -0.1 -0.1 -0.1 EU -0.3 -0.1 3.6 -0.3 -0.5 -0.3 -0.2 -0.3 -0.2 -0.2 -0.2 US 5.3 -0.4 0.0 -0.6 -0.9 -0.6 -0.6 -0.6 -0.5 -0.5 -0.5