Food Systems and the Ecological Footprint of the EU-27: 2004-2014

The study addresses how EU-27 food systems depend on biological resources and contribute to carbon emissions, and how these patterns evolved from 2004 to 2014. Framed by global ecological overshoot and the high share of food systems in global GHG emissions (about one-third), the paper situates the EU’s Green Deal and Farm to Fork Strategy as key policy contexts. It explores whether current consumption patterns and trade dependencies keep EU food demand within regional biocapacity and compatible with decarbonization targets. The purpose is to quantify the ecological footprint (EF) and its food component (food footprint, FF), trace sourcing dependencies via trade, and assess dietary and waste-related drivers, thereby informing policy for sustainable and resilient EU food systems and strategic autonomy.

Background research highlights: (1) food systems account for roughly 34–37% of global GHG emissions; (2) current diets in industrialized countries drive transgression of planetary limits and constitute a large share of national environmental footprints; (3) projected increases in food demand to 2050 risk escalating land, water, and nutrient pressures; (4) within the EU, assessments suggest that supply-side measures alone are insufficient to align with climate goals, and that dietary shifts can markedly reduce carbon, water, and land footprints; (5) concerns exist over potential leakage effects from EU sustainability policies; and (6) previous studies examined footprints at national/city levels or focused on single indicators, but lacked a comprehensive, EU-27-wide, multi-year analysis combining land appropriation and carbon emissions. This study addresses those gaps by applying an EF-extended MRIO approach across EU-27 over a decade.

Design: A consumer-based ecological footprint (EF) assessment extended with a multi-regional input–output (MRIO) model (EF-MRIO) to quantify the EU-27’s food footprint (FF) and sourcing dependencies for 2004, 2007, 2011, and 2014. Scope and metrics: EF quantifies human demand for biologically productive land/sea area (global hectares, gha) across six asset types: cropland, grazing land, forest land, fishing grounds, built-up land, and carbon-uptake land (for CO2 sequestration). Biocapacity (BC) quantifies regenerative capacity using country-specific areas, yield factors, and equivalence factors. MRIO framework: The Global Trade Analysis Project (GTAP) version 10 MRIO (141 regions, 65 sectors) was used to attribute EF embedded in global supply chains to final consumption. Environmental extension matrices allocate EF of production to GTAP sectors (cropland to sectors 1–8; grazing 9–12; forest 13; fishing 14; carbon-uptake allocated by sectoral emissions; built-up distributed by sectoral value added). Final demand is mapped to UN COICOP household consumption categories to derive the FF (household food consumption EF). Consumption accounting: EF of consumption is computed by combining sectoral EF intensities with the Leontief inverse to attribute direct and indirect EF across supply chains to household final demand. Results are reported by COICOP food macro-categories (e.g., bread and cereals; meat; fish and seafood; milk, cheese and eggs; oils and fats; fruits; vegetables; sugars and confectionery; beverages; food products n.e.c.). Dietary EF intensities: Country-level FF intensities (gha per kg and gha per 1,000 kcal) were calculated by dividing per capita FF by FAO Food Balance Sheets (FBS) per capita food supply (kg and kcal). A concordance was developed to allocate FBS items to COICOP categories, using CPC v2.1 and world crop trade shares to split composite items where necessary. Food waste data: Country data for household, food service, and retail levels were taken from UNEP Food Waste Index Report (2021) to estimate potential FF reductions from waste elimination. Temporal coverage: Limited to years available in GTAP 10 (2004, 2007, 2011, 2014). Results focus on EU-27 aggregate and individual Member States. Assumptions/notes: The EF carbon component tracks CO2 sequestration demand only (other GHGs excluded). Ecosystem impacts beyond resource-demand pressures are not quantified. Food-related share of BC supply cannot be precisely isolated due to lack of end-use resolution.

- EU-27 ecological deficit: Average per capita EF fell 20% from 4.34 gha (2004) to 3.47 gha (2014), yet remained above regional BC, which decreased 4% from 2.31 to 2.21 gha per capita. - Food’s role: FF represented 28–31% of total EF and appropriated over half of regional BC. Per capita FF declined from 1.32 gha (2004) to 1.06 gha (2014), remaining about 30% of EF. - Composition: Food is the largest EF category, ahead of personal transportation (21–22%), housing (18–19%), goods and services (14–16% each). Cropland is the dominant asset supporting FF (57%), followed by carbon-uptake land (20–21%), fishing grounds (10–11%), and grazing land (8%). - Country contributions: Five countries account for ~70% of EU-27 FF—Germany 21%, France 15%, Italy 13%, Spain 12%, Poland 8%. Per capita FF highest in Luxembourg (>2 gha), followed by Lithuania, Latvia, Belgium (>1.5 gha). Lowest in Bulgaria, Hungary, and Ireland (<0.6 gha), all below the EU-27 average (1.06 gha in 2014). - Trends 2004–2014: Per capita FF decreased in most countries (notably Cyprus −36%, Czech Republic −35%, Hungary/Luxembourg/Romania −30%), but increased in Latvia (+41%), Lithuania (+24%), Malta (+4%). FF share of total EF remained ~30% with few country exceptions. - Diet drivers: Bread and cereals, meat, and fish/seafood contribute ~49% of FF but only 27% of food supply by weight. Animal-based items have higher EF intensities than plant-based foods. High consumption of bovine meat and high-trophic fish elevates FF. - Potential reductions: Halving beef intake (−13 kg per capita year) and replacing kcal with poultry/pig or pulses could reduce per capita FF by ~6–7% (from 1.06 to ~0.99 gha). Eliminating retail/food service/household food waste could lower FF by an average 13% across EU-27 (range 9% in Austria/Belgium to 19% in Malta). - Trade dependence: In 2014, 24% of EU-27 FF relied on BC from non-EU countries; intra-EU reliance increased (FF sourced within EU from 74% in 2004 to 76% in 2014), with increased intra-regional trade (ROE share up from 22% to 30%) and slight decrease in ROW share (26% to 24%). Average reliance on foreign BC rose at the country level from 48% (2004) to 54% (2014), exceeding two-thirds of FF in 10 countries (e.g., Austria, Belgium, Cyprus, Estonia, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Slovenia). - Notable cases: Netherlands met only 22% of FF from domestic BC in 2014 (down from 26%), increasing ROE and ROW shares; top BC partners include Brazil, Germany, France. France and Germany relied on imports for 49% and 58% of FF, respectively, in 2014. - Policy-related drivers: Declines in EF components relate to EU decarbonization policies and shifts in meat consumption (lower red meat, higher white meat).

The analysis shows that while the EU-27 reduced both total EF and FF between 2004 and 2014, the region continued to operate in ecological deficit, with food consumption remaining the largest EF component. The EF-MRIO results clarify the EU’s food-system resource dependencies: substantial reliance on cropland and carbon-uptake services and increasing dependence on biocapacity outside national borders, though more within EU than extra-EU. These findings directly address the research question by quantifying the resource use and embedded emissions of EU diets and tracing trade-related externalization of pressures. They indicate that domestic supply-side improvements alone cannot deliver sustainability targets; demand-side actions (dietary shifts to plant-rich patterns, reducing beef and high-trophic fish, and curbing food waste) are essential. The observed increase in intra-EU sourcing aligns with EU Strategic Autonomy aims, but the persistent ~25% extra-EU BC reliance highlights exposure to global supply risks and potential offshoring of environmental impacts. The results reinforce policy needs under the Green Deal and Farm to Fork strategies—linking production standards with consumption measures, procurement, education, and trade policies that internalize environmental costs and reduce external dependency.

This study provides the first comprehensive EF-MRIO assessment of the EU-27 food footprint across 2004–2014, quantifying land appropriation and carbon-uptake demand, identifying diet categories driving impacts, and mapping sourcing dependencies via trade. Key contributions include evidence of (1) significant but insufficient reductions in EF/FF amid ongoing ecological deficit; (2) food consumption as the dominant EF component; (3) increased intra-EU sourcing but continued quarter-scale reliance on extra-EU biocapacity; and (4) material mitigation potentials from dietary shifts (particularly reduced beef) and food waste elimination. Future research should: update analyses beyond 2014 to reflect recent shocks (COVID-19, Ukraine conflict); develop dynamic scenarios quantifying savings from dietary transitions and procurement reforms; expand to multi-model MRIO comparisons to test sensitivity; incorporate non-CO2 GHGs and broader environmental pressures; and further disaggregate food waste impacts to distinguish prevention versus dietary effects.

- Temporal scope: Analysis limited to 2004–2014 due to GTAP v10 data availability; results may not reflect post-2014 dynamics (e.g., COVID-19, Ukraine war). - EF carbon component: Tracks CO2 sequestration demand only; excludes other GHGs (e.g., methane, nitrous oxide), underestimating pressures from livestock and fertilizers. - Pressure vs impacts: EF measures resource-demand pressures, not ecological outcomes (e.g., soil degradation, overfishing), and considers flows not resource stocks; many ecosystem services (regulating, cultural) are not captured. - Food waste treatment: Embedded within national FF values but not isolated, hindering precise attribution of reductions from waste prevention vs dietary change. - Dietary shift modelling: Potential savings from diet changes are not comprehensively quantified; dynamic modelling is needed. - MRIO model choice: Results depend on the chosen MRIO (GTAP v10); different MRIOs can yield varying footprint estimates. - BC allocation: Lack of detailed end-use data prevents isolating food-only BC supply, likely underestimating food pressures on ecosystems. - Data access: Some datasets (GTAP) are licensed/restricted, constraining reproducibility of full pipelines (though methods are described).