Individual Farm Model for Common Agricultural Policy Analysis
IFM-CAP is a micro model designed for the ex-ante economic and environmental assessment of the medium-term adaptation of individual farmers to policy and market changes. IFM-CAP was developed by JRC in close cooperation with DG AGRI starting from 2013 for the purpose to improve the quality of agricultural policy assessment upon existing aggregate (regional, farm-group, …) models and to assess distributional effects of policies over the EU farm population. Rather than providing forecasts or projections, the model aims to generate policy scenarios, or ‘what if’ analyses. It simulates how a given scenario, for example, a change in prices, farm resources or environmental and agricultural policy, might affect a set of performance indicators important to decision makers and stakeholders.
IFM-CAP is a comparative static positive mathematical programming model applied to each individual farm from the Farm Accountancy Data Network (FADN) to guarantee the highest possible representativeness of the EU agricultural sector. Farmers are assumed maximizing their expected utility at given yields, product prices and CAP subsidies, subject to resource endowments and policy constraints. The main strengths and capabilities of the model include the possibility to conduct a flexible assessment of a wide range of farm-specific policies and to capture the full heterogeneity of EU commercial farms in terms of policy representation and impacts (e.g. small versus big farms).
IFM-CAP can be applied for ex-ante economic and environmental impact assessment of agricultural and environmental policies at micro (farm) level. For example, IFM-CAP was applied to support the DG AGRI Impact Assessment accompanying the proposal for the CAP post 2020 (SWD/2018/301).
- Licence type:
- Non-Free Software licence
details on model structure and approach
IFM-CAP is a static positive mathematical programming model applied to each individual FADN (Farm Accountancy Data Network) farm (83 292 farms). It assumes that farmers maximise their expected utility at given yields, product prices and CAP subsidies, subject to resource endowments (arable land, grassland and feed) and policy constraints, such as CAP greening restrictions. Farmers’ expected utility is defined following the mean-variance approach with a constant absolute risk aversion specification. Following this approach, expected utility is defined as expected income and the associated income variance. Effectively, it is assumed that farmers select a production plan that minimizes the variance in income caused by a set of stochastic variables for a given expected income level.
Farmer’s expected income is defined as the sum of expected gross margins minus a non-linear (quadratic) activity-specific function. The gross margin is the total revenue including sales from agricultural products and direct payments (coupled and decoupled payments) minus the accounting variable costs of production activities. Total revenue is calculated using expected prices and yields assuming adaptive expectations (based on the previous three observations with declining weights). The accounting costs include the costs of seeds, fertilisers and soil improvers, crop protection, feeding and other specific costs. The quadratic activity-specific function is a behavioural function introduced to calibrate the farm model to an observed base-year situation, as usually done in positive programming models. This function intends to capture the effects of factors that are not explicitly included in the model, such as farmers’ perceived costs of capital and labour, or model misspecifications.
Regarding income variance, most of the models in the literature incorporate uncertainty in the gross margin per unit of activity or in the revenues per unit of activity. The former models assume that prices, yields and costs are stochastic. The latter models either consider that costs are non-random because they are assumed to be known when decisions are made, or are less stochastic than revenues from the farmer’s perspective. Thus, the variance in the gross margin can be approximated by the variance in revenues. In the IFM-CAP framework, the second approach is applied by considering uncertainty only in prices and yields (i.e. revenues) but without differentiating between sources of uncertainty.
IFM-CAP is calibrated for the base year 2012 using cross-sectional analysis (i.e. multiple observations) and Highest Posterior Density (HPD) approach with prior information on regional supply elasticities and dual values of resources (e.g. land rental prices). The calibration to the exogenous supply elasticities is performed in a non-myopic way by taking into account the effects of changing dual values on the simulation response.
The primary data source used to parameterize and calibrate IFM-CAP is individual farm-level data available from the Farm Accountancy Data Network (FADN) database complemented by other external EU-wide data sources such as Farm Structure Survey (FSS), CAPRI database and Eurostat. All farms represented in the FADN sample for the year 2012 (83 292 farms) are included in the model. However, to obtain expected income, past observations (2007–2012) on yields, prices and input costs for these farms are also used for model parameterisation and calibration.
One needs to be aware when applying IFM-CAP that the policy simulations obviously reflect the assumptions in the model. First, the current version of IFM-CAP assumes a fixed farms structure, implying that land can be reallocated only within farms in response to the simulated policy changes. A second potential caveat of the model is that market feedback effects (output price changes) are not taken into account. Third, certain crops are defined in the model as an aggregation of a set of individual crops (e.g. ‘other cereals’). Fourth, FADN includes only commercial farms; small non-commercial farms are underrepresented in the database. A careful analysis of each of these limitations of the current version of IFM-CAP model is needed to be taken into account when analyzing the simulation results.
The following list includes the key data inputs used in the IFM-CAP model:
- Utilised Agricultural Area (FADN)
- Arable and grassland (FADN)
- Set of crop and livestock activities (FADN)
- Yields, Prices and Subsidies (FADN)
- Observed activity levels (hectares of crop area and number of livestock) (FADN)
- Farm level feed costs (FADN)
- Farm weighting factor (FADN)
- Land and milk quota rental prices (FADN)
- Prices and yields for fodder crops at MS level(FADN and CAPRI)
- Feed prices at MS level (CAPRI)
- Feed nutrient content (CAPRI)
- Nutrient requirement of animal activities (NRC , IPCC , LfL , CAPRI)
- Price and yield trends(CAPRI)
- Elasticities for feed demand at NUTS2 level (CAPRI)
- Supply elasticities for livestock activities (CAPRI)
- Supply elasticities for crops at NUTS2 level (Jansson and Heckelei, 2011)
- Carcass weights (Eurostat)
- Prices of live animals (Eurostat)
- Out-of quota prices for sugarbeet(Agrosynergie, 2011)
- MS sugarbeet in-quota production (DG-AGRI,2014)
- In- quota prices for sugar beet (Agrosynergie, 2011)
- Soil erosion cover-management factors (Panagos et al., 2015)
The main outputs/indicators generated by IFM-CAP for a specific policy scenario are the following:
- Land allocation/crop area (ha)
- Herd size/animal number (heads)
- Livestock density (LU/ha)
- Share of arable land in Utilized Agricultural Area
- Share of grassland in Utilized Agricultural Area
- Land use change (ha)
- Agricultural production (Tons)
- Intermediate Input use (Tons)
- Agricultural output (€)
- CAP first pillar subsidies (€)
- CAP second pillar subsidies (€)
- Intermediate input costs (€)
- Variable costs (€)
- Total costs (€)
- Gross farm income (€)
- Net Farm Income (€)
- Biodiversity index
- Soil erosion
model spatial-temporal resolution and extent
|Spatial Extent/Country Coverage|
EU Member states 27
EU-wide model covering the EU agricultural sector
Medium-term (5 to 15 years)
Used for medium-term comparative analyses. The time horizon (i.e. Baseline) for running simulation is 2025 and 2030 depending on the policy scenario.
Static model, simulations are done for a single time point (year), without any intermediate steps.