The European agro-economic model AROPAj
Résumé
This publication presents the European agro-economic AROPAj model. Throughout recent years the model has been adapted to suit FADN data and the expansion of the European Union (EU ). The revised model, which is presented here, is based on the same principles,
modularity, and generic conceptions of the previous renditions. The strength of the model lies in the evolution of computing capacities and facilities, especially when new scientific problems arise.
The starting premise of the model resulted from the combination of different factors, namely new policy tools announced to be implemented by the changing Common Agricultural Policy (CAP), the need to account for a wide diversity of farming systems, and the changing geographical area following the EU enlargement. In complement of econometric approaches relying on past information, the model introduces new mathematical programming (MP) approaches using the most relevant data.
The Farm Accountancy Data Network (FADN ) is considered as a permanent, robust and representative source of information, useful for MP models delineated at the farm level. The information in the FADN database consists of surveys of individual farm samples. The database provides individual information mainly accounting of accountancy and financial or economic variables. The AROPAj model is also capable of using data on physical variables such as livestock, production, and land. Nevertheless, a lot of AROPAj parameters are not directly sourced by raw FADN data. For example, the model makes use of other data sources, including expert reports and books, such as livestock feeding. Other sources utilized by some AROPAj modules are inter alia data concerning land use and land cover (Land Use/Cover Area frame Statistical Survey (LUCAS), CORINE Land Cover (CLC)), the European Soil Database provided by the European Joint Research Centre (JRC), the Digital Elevation Model (DEM ), and the Climate information provided by the European MARS JRC Unit. These variables are grouped by means of a clustering method to form a group of similar farms that present similar characteristics and thus can form a unit of modeling for the AROPAj model: the farm type (also denominated by farm group). The overall model is based on farm-groups, designed on a regional scale, for all of the EU.
In large part, the AROPAj model originates from modeling tools aiming at CAP impact assessments. Mathematical programming models are effective in this way, especially when the implementation of new policy tools in economic models has to rely on non-time dependent data, unlike time-inferred econometric models. Due to the wide and complex diversity of CAP tools renewed throughout decades, the CAP block size quickly increased. The generic modeling approach embraces the wide variety of European farming systems which are subsumed by the continually-growing EU. It involves a multi-scale analysis from the farm type level to the EU
level, accounting for realistic discontinuities at the farm level.
The model is also designed to assess global change impacts and analyze policy while accounting for the complexity of interactions between agriculture and the environment. In this regard, it is a tool used for the analysis of a large range of economic and environmental domains, such as greenhouse gas (GHG) emissions and abatement, nitrogen sourced pollutions, biomass provision from perennial crops, ozone impacts and adaptation to climate change. Other collateral outputs may exist when providing farm-based estimates and data related to physical parameters.
The interest of linking the AROPAj economic model with biophysical models relies on the possibility of accounting for environmental (physical impacts such as soil and climate) and technical variable impacts on the production functions, as well as on the agricultural system output relevant to environmental concerns. The general goal is to bring to light the relationship between crop production inputs (e.g. fertilizer applications) and crop production outputs (e.g. yields). The potential relation between the economic and biophysical models is to be found in the management practices of the farmer, such as irrigation, application of pesticides, and fertilizer.
Some refinements are required when the farm group’s optimal solution is made of non-marketed crop output (or of non-marketed nitrogen input). In other words, this occurs when, at the optimum, the crop production is entirely re-used on-farm (e.g. for livestock feeding), or when no N -input comes from the market (e.g. manure sourced). In this case, market prices have to be replaced by relevant shadow prices, which are part of the optimal solution. In practice, the problem is solved by iterated AROPAj runs with shadow prices replacing the prices used at the previous iteration for the yield function derivative. This process of convergence is fast in most AROPAj LP’s, except for a few cases where solutions have periodic forms over iterations, with slight amplitudes.
The model is running using the GNU software and applications, except for the marketed optimization solving tool (GAMS). Year after year, the choice has been made to promote the use of GNU software for AROPAj improvement, not only regarding the financial cost, but
mainly to facilitate the maintenance and the evolution of our programs and applications. The model also produces some interesting spatially disaggregated outputs in line with multi-scale environmental problems. The computing and algorithm facilities embedded in the modeling process allow new improvements and new model exploitation exercises.
The European agro-economic AROPAj model is built of a set of linear programming models, aiming at a good representation of European farming systems in term of geography and farm type diversity as well. As a result, the model aims at being a tool and a reference, which benefits a significant numbers of PhDs and MSc students, and contributes to many published peer-reviewed papers (to which this document refers).
Domaines
Economies et finances| Origine | Fichiers produits par l'(les) auteur(s) |
|---|
