Elnaz Nejatiyan Pour1, Nicola Gallai2, Georgios Kleftodimos3, Peter Zander1, María Rosa Mosquera-Losada4, Seyed-Ali Hosseini-Yekani *1
1- Leibniz Centre for Agricultural Landscape Research (ZALF)
2- ENSFEA UMR LEREPS
3- CIHEAM-IAMM
4- University of Santiago de Compostela
2- ENSFEA UMR LEREPS
3- CIHEAM-IAMM
4- University of Santiago de Compostela
Abstract: (4 Views)
Introduction and Objectives: Pollinator decline is an important ecological threat to European agri-food systems because animal-mediated pollination contributes to the productivity, quality, and stability of many high-value crops. The economic importance of this ecosystem service extends beyond the farm level because pollination-dependent agricultural products are connected to input markets, food-processing industries, household consumption, and other sectors of the economy. Many previous studies have focused on biophysical, agronomic, or direct economic valuation at crop or farm level. These approaches are essential for demonstrating the importance of pollination, but they do not fully represent price adjustments, substitution among commodities and production factors, or feedbacks across sectors. This study therefore develops a transparent and replicable computable general equilibrium (CGE) framework to evaluate the economy-wide effects of pollination-service shocks and to compare two types of policy intervention in European agri-food systems. The main methodological contribution is to represent pollination services explicitly as a distinct primary input in the production structure of pollination-dependent crops, rather than approximating pollinator decline only as a generic productivity shock.
Materials and Methods: The analysis uses a static, single-country CGE model calibrated to the 2015 Bio-Social Accounting Matrix (BioSAM) for Italy. To align the database with the research objective, production and commodity accounts were aggregated into five groups: pollination-dependent crops, pollination-independent crops, livestock and animal products, food-processing industries, and other non-agricultural activities. Primary factors include labour, capital, and pollination services (POL), while the institutional structure includes households, enterprises, government, and the rest of the world. Model parameters were calibrated so that the benchmark equilibrium reproduces the flows recorded in the 2015 SAM. Crop-specific pollination-dependence coefficients were based on the classification of Klein et al. (2007). For aggregated accounts containing crops with different dependence levels, harvested-area shares were used in the aggregation procedure. The corresponding pollination-service share was then separated from the value-added block of dependent activities and recorded as the POL factor, while remaining value added was allocated between labour and capital in proportion to their initial shares. This reallocation preserves total value added and the balance of the SAM. In the aggregated model, the calibrated POL share in value added of pollination-dependent crops is 0.40. Commodity markets follow the Armington assumption for imperfect substitution between domestic and imported goods and a constant-elasticity-of-transformation (CET) structure for allocation between domestic sales and exports. Product-level trade elasticities were obtained from the GTAP Database and averaged according to the product composition of each aggregated account; the final Armington and CET elasticities used in the model are 0.70 and 2.00, respectively. Under the macroeconomic closure, labour and capital supplies are fixed and fully employed, the exchange rate is fixed, and foreign savings adjust endogenously to maintain the external balance. The exploratory simulations include 10% and 30% reductions in pollination effectiveness, a 30% reduction combined with targeted production support, and a 30% reduction combined with partial ecosystem restoration. The 10% and 30% shocks are stylized exploratory intensities and should not be interpreted as empirical estimates or forecasts of future pollination decline.
Results: The results show that reductions in pollination services produce substantial and uneven sectoral effects, with the largest losses concentrated in pollination-dependent crops. Under the 10% decline scenario, output of pollination-dependent crops and food processing decreases by approximately 4.7% and 3.5%, respectively. When the shock increases to 30%, the corresponding output losses reach approximately 14.5% and 11.5%. Under the targeted production-support scenario, the declines in these two groups are moderated to approximately 11.1% and 8.7%, while under partial ecosystem restoration they are limited to approximately 7.1% and 5.4%. Output changes in the rest of the economy remain very small across all scenarios. Price effects also increase with shock intensity. Under the 10% decline scenario, producer prices rise by 3.53%, food prices by 0.06%, and real income of the representative household falls by 0.08%. Under the 30% decline scenario, producer and food prices increase by 12.56% and 0.20%, respectively, while real household income decreases by 0.28%. In the production-support scenario, producer prices rise by 9.19%, food prices by 0.15%, and real household income declines by 0.20%. Under partial restoration, the corresponding changes are +5.50%, +0.09%, and -0.12%. Wages and returns to capital also decline in the pollination-shock scenarios. The model-implied return to the POL factor falls by 1.02% and 2.69% under the 10% and 30% declines, rises by 3.95% under the support scenario, and falls by 1.50% under restoration. Given the fixed-exchange-rate closure, foreign savings increase by 2.94% and 11.45% under the 10% and 30% shocks and by 8.05% and 4.69% under support and restoration, respectively. These values represent closure-specific external-account adjustments and should not be interpreted as direct changes in imports or exports. Because the model contains a single representative household, the welfare results cannot be used to infer differences across income or social groups.
Conclusion: The findings show that pollination services are not only ecological assets but also economically important production inputs whose disruption can propagate through production linkages, relative prices, and factor incomes. Both policy interventions mitigate part of the severe shock. Under the parameterization used in this study, partial restoration produces a larger moderation of output losses, price pressures, and the decline in real household income than targeted production support. However, because the two interventions do not have identical costs or designs, this result should not be interpreted as a cost-effectiveness ranking. From a policy perspective, the results support considering short-run economic support together with measures aimed at protecting and restoring pollination services, while recognizing that the appropriate policy mix depends on implementation costs, budget constraints, and the degree of ecosystem-service recovery. Because the model is static, the reported results represent comparative-equilibrium differences across scenarios rather than temporal pathways of decline or recovery. Future extensions using a dynamic CGE framework and disaggregated households could examine adjustment paths, dynamic costs, and distributional effects.
Materials and Methods: The analysis uses a static, single-country CGE model calibrated to the 2015 Bio-Social Accounting Matrix (BioSAM) for Italy. To align the database with the research objective, production and commodity accounts were aggregated into five groups: pollination-dependent crops, pollination-independent crops, livestock and animal products, food-processing industries, and other non-agricultural activities. Primary factors include labour, capital, and pollination services (POL), while the institutional structure includes households, enterprises, government, and the rest of the world. Model parameters were calibrated so that the benchmark equilibrium reproduces the flows recorded in the 2015 SAM. Crop-specific pollination-dependence coefficients were based on the classification of Klein et al. (2007). For aggregated accounts containing crops with different dependence levels, harvested-area shares were used in the aggregation procedure. The corresponding pollination-service share was then separated from the value-added block of dependent activities and recorded as the POL factor, while remaining value added was allocated between labour and capital in proportion to their initial shares. This reallocation preserves total value added and the balance of the SAM. In the aggregated model, the calibrated POL share in value added of pollination-dependent crops is 0.40. Commodity markets follow the Armington assumption for imperfect substitution between domestic and imported goods and a constant-elasticity-of-transformation (CET) structure for allocation between domestic sales and exports. Product-level trade elasticities were obtained from the GTAP Database and averaged according to the product composition of each aggregated account; the final Armington and CET elasticities used in the model are 0.70 and 2.00, respectively. Under the macroeconomic closure, labour and capital supplies are fixed and fully employed, the exchange rate is fixed, and foreign savings adjust endogenously to maintain the external balance. The exploratory simulations include 10% and 30% reductions in pollination effectiveness, a 30% reduction combined with targeted production support, and a 30% reduction combined with partial ecosystem restoration. The 10% and 30% shocks are stylized exploratory intensities and should not be interpreted as empirical estimates or forecasts of future pollination decline.
Results: The results show that reductions in pollination services produce substantial and uneven sectoral effects, with the largest losses concentrated in pollination-dependent crops. Under the 10% decline scenario, output of pollination-dependent crops and food processing decreases by approximately 4.7% and 3.5%, respectively. When the shock increases to 30%, the corresponding output losses reach approximately 14.5% and 11.5%. Under the targeted production-support scenario, the declines in these two groups are moderated to approximately 11.1% and 8.7%, while under partial ecosystem restoration they are limited to approximately 7.1% and 5.4%. Output changes in the rest of the economy remain very small across all scenarios. Price effects also increase with shock intensity. Under the 10% decline scenario, producer prices rise by 3.53%, food prices by 0.06%, and real income of the representative household falls by 0.08%. Under the 30% decline scenario, producer and food prices increase by 12.56% and 0.20%, respectively, while real household income decreases by 0.28%. In the production-support scenario, producer prices rise by 9.19%, food prices by 0.15%, and real household income declines by 0.20%. Under partial restoration, the corresponding changes are +5.50%, +0.09%, and -0.12%. Wages and returns to capital also decline in the pollination-shock scenarios. The model-implied return to the POL factor falls by 1.02% and 2.69% under the 10% and 30% declines, rises by 3.95% under the support scenario, and falls by 1.50% under restoration. Given the fixed-exchange-rate closure, foreign savings increase by 2.94% and 11.45% under the 10% and 30% shocks and by 8.05% and 4.69% under support and restoration, respectively. These values represent closure-specific external-account adjustments and should not be interpreted as direct changes in imports or exports. Because the model contains a single representative household, the welfare results cannot be used to infer differences across income or social groups.
Conclusion: The findings show that pollination services are not only ecological assets but also economically important production inputs whose disruption can propagate through production linkages, relative prices, and factor incomes. Both policy interventions mitigate part of the severe shock. Under the parameterization used in this study, partial restoration produces a larger moderation of output losses, price pressures, and the decline in real household income than targeted production support. However, because the two interventions do not have identical costs or designs, this result should not be interpreted as a cost-effectiveness ranking. From a policy perspective, the results support considering short-run economic support together with measures aimed at protecting and restoring pollination services, while recognizing that the appropriate policy mix depends on implementation costs, budget constraints, and the degree of ecosystem-service recovery. Because the model is static, the reported results represent comparative-equilibrium differences across scenarios rather than temporal pathways of decline or recovery. Future extensions using a dynamic CGE framework and disaggregated households could examine adjustment paths, dynamic costs, and distributional effects.
Keywords: Computable General Equilibrium (CGE) modelling, Pollination services, Ecosystem service shocks, Agri-food systems, Resilience
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
