Methods

surpluses are especially present in regions of high animal density (Svanbäck et al. 2019). The European Commission’s proposal on the CAP reform states that the policy framework shall more strongly consider ‘the need to improve farms sustainability, and in particular the nutrients management’ (European Commission 2018, paragraph 22) as well as ‘the response of EU agriculture to societal demands on […] animal welfare’ (European Commission 2018, specific objectives (i)). Restricting animal density and nitrogen application can potentially become part of the future EU agricultural policy under animal welfare and environmental considerations.

exchange of production quantities and market price changes. In this global model, consumers, producers and traders interact as economic agents based on microeconomic theory. Trade flows are modelled in a two-stage demand system based on the “Armington (1969) assumption” that differentiates between domestic sales and imports as well as between imports of different origins. The underlying reasoning in the CAPRI implementation is that consumers substitute less easily between domestic and imported goods than they do between imported goods of different origins. In addition to effects on quantities and prices, a number of environmental indicators (e.g., nutrient surpluses and greenhouse gas emissions from the agricultural sector) are also calculated in the modelling system.

4.3.2 Scenario design

The chosen reference scenario is based on the “Agricultural Outlook” of the European Commission (2016). In this scenario, the current CAP is extended until 2030. Technological progress, and population and economic growth are projected based on trend assumptions. As this scenario is based on the currently implemented EU agricultural policy, it can be interpreted as a

“business-as-usual” (BAU) scenario.

While direct payments to farmers are organized within the first pillar of the CAP, the second pillar is designed to support rural areas within the EU.

Second pillar measures are modeled in line with the actual regulations covering “Less Favoured Area” payments, agricultural-environmental measures, or “Natura 2000” support for biodiversity protection. In the first alternative policy scenario, we analyze a reduction of first pillar direct payments by 50 per cent (DP50) based on the respective amount paid in the BAU scenario. In this scenario, the capped direct payments drop completely out of the CAP budget. The reduction is implemented as a cut in all measures in the first pillar of the CAP, including decoupled direct payments and voluntary coupled support. While decoupled payments are independent of production levels, some degree of coupling remains because land receiving payments is supposed to be kept in good agricultural and environmental condition and must not be abandoned.

Cutting Pillar I payments in half is a rather unlikely setting; still reductions in the CAP budget are part of the EU reform debate. Therefore, this potentially extreme case is tested to assess the implications of potential CAP budget cuts in the EU for agricultural trade and environmental sustainability.

Our second scenario is designed as a transfer of the budget freed-up by cutting the payments previously related to CAP Pillar I to measures with a focus on extensive crop production in Pillar II (DPTRANS). In the implementation, extensive crop production is represented as a production technology requiring fewer inputs that is, however, as well reflected in lower yields. Also, the shift of some Pillar I payments for a broad range of agricultural activities to financial support of mainly crop-producing activities induces some changes in the agricultural sector.

The scenario is inspired by the proposal of allocating 30 per cent of the Pillar I payments to schemes for organic farming, permanent grasslands, or marginal areas (European Commission 2018). In the discussion on the future CAP, Matthews (2018) describes a planned transfer of 15 per cent of the Pillar I national ceilings to environmental and climate measures in the second pillar. Our scenario exceeds these suggestions and the probable CAP changes to emphasize the potential of such a transfer.

Areas of high animal density are hotspots for nitrogen surpluses and related soils and water pollution (Jørgensen et al. 2018). To account for regional heterogeneity regarding nutrient balances, we restrict maximum animal density in a further scenario (LSMAX) to the respective local soil nitrogen needs in the BAU scenario. In detail, we simulate this scenario by dividing the regional nitrogen need per hectare taken from the CAPRI nutrient balances by the regional excretion per livestock unit in a region based on the BAU scenario to define the maximum livestock density per hectare. In the regional programming models, this upper bound is implemented as an inevitable constraint.

In this way, we prevent a nutrient undersupply of the soil and related strong negative consequences for yields and plant productivity (Csathó and Radimszky 2009). The shock is attenuated in areas with low soil nitrogen needs by implementing a minimum boundary of 0.6 livestock units per ha.

This limit lies within the boundaries that Buckwell and Nadeau (2018) describe as sustainable animal density for ruminants. In Appendix A, we provide an overview of livestock densities before and after the restriction across EU regions (Table 4.9 in Appendix A).

While EU nitrogen surpluses have generally declined in hotspot areas, strong surpluses persist and the overall surplus level in the EU remains high by international comparison (van Grinsven et al. 2012; Potter et al. 2010). In the CAPRI modelling system, we simulate an enforced Nitrates Directive by imposing soil nitrogen surplus limits of 50 kg N ha^-1 a^-1 (NITR). Fertilizer applications influence the nutrient balances in the model and are configured in a way that the soil nitrogen surplus must not exceed the stricter limit. The resulting reduction in nitrogen surpluses varies by region and its nitrate vulnerability status. For some regions, nitrogen surpluses even reduce to one-eighth of the surplus level in the BAU scenario. This enforcement is implemented on top of other nitrate directive components taken from existing regulations without further adjustment in our scenario design (e.g.

a 170 kg N ha^-1 a^-1 manure application limit, regional maximum fertilization specifications based on EU member state regulations).

Furthermore, we assess the restriction of animal density and nitrogen application in a combined approach (NCOMBI). Practically, we combine the scenarios by simulating the nitrogen surplus limit of 50 kg N ha^-1 a^-1 and the livestock density restriction in one run. Since the livestock density restriction is designed based on livestock numbers and nutrient balances from the BAU simulation, the specification of the constraint is not affected by changes in the actual nitrogen balances of the current scenario run.

However, the nitrogen surplus as such (even though not the implemented policy restriction) can be affected by the livestock restriction. Also, in the scenarios NITR and NCOMBI, the imposed constraints on nitrogen surplus may contribute to lower livestock densities. An overview on the scenarios used in this study is provided in Table 4.1.

Table 4.1 Scenario overview

Scenario group Acronym Description Business-as-usual BAU Reference scenario Adjustments of direct

payments (DP)

DP50 CAP Pillar I payments reduced by 50%

DPTRANS Budget reduced in DP50 transferred to CAP Pillar II

Restrictions of animal density and nitrogen application

LSMAX Livestock density restriction NITR Surplus nitrogen limitation

NCOMBI Combination of LSMAX and NITR

4.3.3 Indicators

In the scenario assessment, we focus on relevant impacts on EU–Africa trade flows. For the reference scenario, agricultural product trade flows are analyzed for the EU and the African model regions in 2030. Policy scenario impacts are assessed on the basis of changes in consumer and producer prices, and production, consumption, import and export quantities. Potential implications for welfare and food security are pointed out, although in the light of limited model representation. Substituting trade flows to Africa from other countries are considered in this analysis as well. We also investigate changes in land-use, nitrogen surpluses and agricultural greenhouse gas emissions given that the policy changes simulated aim at increased environmental sustainability.