The European Green Deal improves the sustainability of food systems but has uneven economic impacts on consumers and farmers

The European Union’s food system is currently unsustainable, contributing about 30% of regional greenhouse gas emissions, driving biodiversity loss, wasting resources, and fostering unhealthy diets. Prior analyses indicate that achieving sustainable and healthy food systems requires simultaneous changes in farming practices, loss reduction along the supply chain, and dietary shifts away from animal-based foods. Motivated by the European Green Deal (EGD), Farm to Fork, and Biodiversity strategies, this study asks: what are the market (prices, production, consumption, trade, farmer revenues, consumer expenditures) and non-market (GHG emissions, biodiversity, nutrition) impacts of three key levers—(1) agro-ecological practices via reduced chemical inputs and more high-diversity landscape features, (2) reduced post-harvest food losses, and (3) shifts toward healthier diets with fewer animal-based products—when applied alone and jointly in the EU-27? The purpose is to provide a comprehensive assessment, including both environmental and nutrition indicators, and to clarify distributional economic effects for consumers and different producer groups.

Global and European studies have shown the importance of combining supply and demand-side measures. Springmann et al. (2018) linked global food demand and production using a reformulated IMPACT partial equilibrium model and found that combining changes in farm practices, loss reduction, and dietary shifts toward more plant-based foods is needed to limit environmental pressures. At the European scale, Clora et al. (2021) used a modified GTAP-E to compare supply-side mitigation (extensification vs intensification) under healthy diet baselines, finding extensification reduces agricultural GHG by ~11% by 2050 while intensification raises them by ~2.5%. Poux and Aubert (TYFA) suggested agro-ecology plus healthy diets could cut EU agricultural GHG by ~40% vs 2010 using a biophysical mass-flow model. Many EGD-focused studies emphasized only farm-level targets (e.g., pesticide/fertilizer cuts) using CAPRI or MAGNET, typically finding lower production, higher prices, and mixed impacts on producer income, with significant carbon leakage partly offsetting domestic GHG gains. The European Commission criticized the supply-only focus for omitting loss reduction and diet shifts. This study fills that gap by integrating post-harvest loss reduction and dietary change, and by evaluating GHG, biodiversity, and nutrition jointly with economic outcomes and explicit producer/consumer price transmission.

The authors develop a synthetic partial equilibrium model of the EU-27 food system calibrated to the average of 2018–2020 (termed 2019). The model aggregates products into three groups on the consumption side: plant products (cereals, oilseeds, protein crops, rice, sugar beets, potatoes; fruits and vegetables excluded), ruminant products (milk, dairy, ruminant meat), and monogastric products (pork, poultry, eggs). On the production side, total agricultural land supply is endogenous and allocated among: (i) food/feed field crops, (ii) forages (permanent and temporary grasslands, non-herbaceous forages) for ruminants, and (iii) high-diversity landscape features (HDLF) for environmental protection. International trade is modeled via net trade, dependent on EU prices (positive with price for plant products where the EU is a net importer at baseline; negative with price for animal products where the EU is a net exporter). Post-harvest losses along the chain and at the consumption stage are explicitly represented, allowing differentiation between production, purchases, and actual consumption. Consumer-producer price gaps are modeled as constant margins calibrated to the base period. Supplies are derived from restricted quadratic profit functions for the three production sectors. Land demands for crop and forage uses are derived from sectoral profit functions. Ruminant and monogastric technologies assume fixed input/output coefficients (Leontief) for concentrated feed purchased from the crop sector. Food demands arise from a separable quadratic utility function yielding a three-equation demand system dependent on final consumer prices; post-harvest losses enter via loss coefficients in these equations. Endogenous variables include producer and consumer prices for the three aggregates, land price, quantities produced, traded (net), purchased, consumed (food, feed), land uses by category, producer revenues, and consumer expenditures. Exogenous variables include HDLF share, chemical input use (pesticides, fertilizers) and veterinary costs (antimicrobials proxy), loss percentages, and consumer preference shifters (to model diet shifts). Calibration uses Eurostat data (2018–2020) for production, land use, yields, trade, and allocation of crop output to food vs feed. Environmental indicators: GHG emissions at farm gate use coefficients per kg by Crenna et al., adjusted per Bellassen et al. to distinguish conventional vs agro-ecological practices and origin (domestic/export vs import). Biodiversity damage uses damage coefficients from Knudsen et al., differentiated by land use (field crops, forages, HDLF), production practices, and origin; forests serve as reference with zero damage. For scenarios with agro-ecology, biodiversity damage coefficients are assumed 10% lower and GHG coefficients 20% lower than conventional (assumption based on expert judgment). Nutrition indicators computed from the three aggregates include total calories, total proteins, animal proteins, plant protein share, fiber, fat, and carbohydrates; decreases in calories, animal proteins, fat, carbohydrates are treated as beneficial while decreased fiber is adverse (within small-change bounds). Scenarios (Table 2): (1) ‘Agro-ecology’ lever—reduced pesticides/fertilizers/antimicrobials and increased HDLF; (2) ‘Food losses’ lever—halving post-harvest losses; (3) Diet lever with two variants: ‘Anim-’ (exogenous −20% demand for ruminant and monogastric products) and ‘Anim-Plant+’ (same −20% animal demand plus iso-protein compensation via +29.5% plant demand); and joint scenarios combining all three levers with and without iso-protein compensation. Sensitivity analyses on elasticities (land supply, supply/demand, trade) are provided in the Archive; agricultural land supply elasticity is set to 0.05 following Tabeau et al. Model solution is implemented in Excel; data and code available via Zenodo.

- Environmental performance: - All three levers improve climate and biodiversity outcomes; strongest gains occur when combined. - ‘Agro-ecology’ alone: total GHG emissions of EU food consumption decrease by 65 Mt CO2-eq (~−4.9% vs base); biodiversity damage −17.6% (trade leakage cancels ~half of domestic gains). - ‘Food losses’ alone: total GHG −76 Mt CO2-eq; biodiversity damage −13%; nearly no change in actual consumed quantities; large reductions in emissions embedded in trade. - Diet shift ‘Anim-’ (−20% animal demand): total GHG −171 Mt CO2-eq; biodiversity damage −27.6%; lower animal proteins (−22.4%), higher plant protein share (0.42→0.48), and decreases in calories, proteins, fat, carbohydrates; fiber almost unchanged. - Diet shift ‘Anim-Plant+’ (iso-protein with +29.5% plant demand): total GHG −103 Mt CO2-eq; biodiversity −11.3%; plant protein share rises to 0.55; fiber +27.2%; however, calories, fat, and carbohydrates increase. - Joint scenario without iso-protein (‘Agro-ecology, food losses, and Anim-’): total GHG −198 Mt CO2-eq (about −20%); biodiversity damage −53.4%. - Joint scenario with iso-protein (‘Agro-ecology, food losses, and Anim-Plant+’): environmental gains remain positive but lower than without compensation due to higher plant production and net imports. - Economic outcomes: - ‘Agro-ecology’ increases producer prices more than quantities fall (inelastic supplies), raising producer revenues for crops (+11.3%), ruminants (+4.5%), monogastrics (+1.8%); consumer expenditures rise modestly (+2.0%) due to limited pass-through (constant margins). Net imports of plant products increase by +32.7 Mt; net exports of animal products fall (ruminants −3.5 Mt; monogastrics −2.1 Mt). - ‘Food losses’: shifts demand left; lowers purchases, production, and prices; producer revenues −9.6%; consumer expenditures −7.0%; net imports of plant products −28.0%; net exports of animal products rise (ruminants +35.8%, monogastrics +33.1%). Actual consumption nearly unchanged due to loss reduction. - ‘Anim-’: strong reductions in purchases/consumption and prices for animal products; domestic production declines are cushioned by much larger net exports (ruminants +109%, monogastrics +92.2%); crop producers face a negative feed effect; producer revenues: ruminants −28.2%, monogastrics −23.6%, crops −8.7%; consumer food expenditures −16.5% (≈−16.8% reported). - ‘Anim-Plant+’: consumer expenditures −3.4%; crop producer revenues +13.9% (food demand effect outweighs reduced feed); livestock revenues drop substantially (about −26.2%); plant net imports increase due to higher prices; animal net exports increase but less than in ‘Anim-’. - Joint scenarios: consumers benefit—expenditures −21.0% (no iso-protein) and −8.8% (iso-protein). Producer revenues: without iso-protein, crops −7.5%, ruminants −28.7%, monogastrics −26.7%; with iso-protein, crops +10.8%, ruminants ~−27.6%, monogastrics ~−26.1%. - Nutrition: - Only diet-shift lever changes nutrition indicators meaningfully; agro-ecology and food-loss levers have negligible nutrition effects. - Iso-protein compensation improves plant protein share and fiber but increases calories, fat, and carbs; no-compensation reduces these intakes but lowers total protein.

The findings show that assessing the European Green Deal’s food-system impacts requires jointly considering supply-side agro-ecological changes, post-harvest loss reductions, and demand-side dietary shifts. Agro-ecology alone modestly improves climate/biodiversity outcomes but increases food prices only slightly given limited pass-through; it also deteriorates the EU food trade balance and induces environmental leakage via trade. Reducing post-harvest losses saves resources throughout the chain, yielding environmental gains, lower consumer spending, and lower producer revenues, with little change in actual consumption. The diet-shift lever is pivotal for nutrition improvements and delivers the largest environmental benefits, but it substantially reduces livestock producer revenues, implying significant sectoral restructuring needs and policy support. Joint implementation enhances climate and biodiversity performance, with consumers generally benefiting economically; however, distributional effects are uneven—livestock sectors lose, while crop producers may gain if increased plant food demand outweighs reduced feed demand (as in iso-protein scenarios). Results are consistent with prior EGD-focused studies on agro-ecology regarding production declines and trade deterioration; the model underscores the importance of land-supply elasticity (allowing limited farmland expansion) and price transmission assumptions. Sensitivity analysis indicates robustness to plausible elasticity ranges, with higher sensitivity to trade parameters. Policy design for achieving these levers (price transmission, anti-leakage measures, targeted support for livestock transition, and actions to reduce losses) will shape final market and welfare outcomes.

Jointly applying the European Green Deal’s three levers—agro-ecological practices, reduced post-harvest losses, and dietary shifts away from animal-based products—substantially improves the climate and biodiversity footprint of EU food consumption. Only dietary change materially enhances nutrition indicators; agro-ecology and loss reduction mainly affect production, trade, and environmental outcomes. Consumers tend to gain via lower food expenditures when all levers are used, but livestock producers face significant revenue losses, necessitating strong transition policies. Crop producers benefit when increased plant food demand (e.g., iso-protein compensation) outweighs reduced feed demand. Future research should (i) refine product aggregation (e.g., separate legumes, ruminant meat vs milk), (ii) incorporate fruits and vegetables and product reformulation to better capture healthy-diet pathways, (iii) model detailed supply-chain price transmission and policy instruments (including anti-leakage measures), and (iv) extend environmental accounting to LULUCF and more comprehensive biodiversity metrics.

- Product aggregation into three groups limits market detail (e.g., no separation of ruminant meat vs milk; cereals vs legumes) and excludes fruits and vegetables, constraining the assessment of fully healthy diets. - Environmental indicators are simplified; biodiversity metrics are partial and rely on assumed coefficient reductions for agro-ecological practices; LULUCF carbon sink/source effects are not included. - Economic assessment omits production cost changes and consumer welfare measures; constant margin assumption for price transmission may not reflect actual vertical relationships. - Diet shifts are modeled as preference changes without explicit policy instruments; alternative policy designs could alter market and welfare outcomes. - Results depend on elasticities and trade parameterization (though sensitivity analysis suggests reasonable robustness); coefficients for agro-ecology GHG (−20%) and biodiversity (−10%) relative to conventional are based on expert judgment in absence of direct estimates. - Estimates of food losses and waste are uncertain; achieving halving of post-harvest losses requires significant behavioral and policy changes.