- John Helming, Wageningen University and Research
- John Helming, Wageningen University and Research
- Paolo Sckokai, Catholic University of the Sacred Heart, Milan (UNICATT)
- Hans van Meijl, Wageningen University and Research
- Ignacio Perez Dominguez, European Commission, Joint Research Centre
- Prof. Simone Severini, DAFNE, Università della Tuscia, Viterbo, Italy
Read the abstracts of the session here:
Policies related to agriculture are expected to address an increasing number of objectives as demanded by society. As a result, agricultural policies, like the European Union (EU) Common Agricultural Policy (CAP), increase their scope to incorporate for example objectives of the Paris climate agreement and the Sustainability Development Goals (SDGs).
A further objective is to significantly simplify and shift the emphasis from compliance and rules towards results and performance of individual farms. The one-size-fits-all approach will be replaced by a more flexible system with greater freedom for EU countries to decide how best to meet the common objectives consistent with specific needs of their own farmers and rural communities. Farm specific measures that will be included in the future CAP are e.g. higher levels of support per hectare for small and medium-sized farms and rewards for farmers for going beyond mandatory agri-environmental and/or climate requirements. Impact assessment tools should include this wider scope and the particular behaviour of individual farmers taking into account initial agronomic, bio-physical, financial, economic and social farm characteristics. In impact evaluations, much more attention has to be paid to the way the CAP is regionally implemented.
For effective and efficient policies the EU needs to take local and regional conditions and policies into account in evaluating its policies at farm, regional, national and global levels. During the last twenty years individual farm, agricultural sector and economy-wide models, hereafter referred to as current models, have been intensively used by the EU Commission for public decision making in, among others, the areas of agriculture, sustainable management of natural resources, ecosystem services and climate change. With the exception of IFM-CAP (Louhichi et al., 2017) the current models were originally developed when the CAP was shifting from market interventions to coupled support. At that time researchers had no or quite limited access to single farm observations. As a result, most current models are not able to deliver impacts for individual farms or local impact as they are specified at higher levels of aggregation. They also struggle to analyse policies specific to the individual farmer or for which interaction between farms and with other agents are crucial for policy outcomes. In many cases, they only provide summary results for the population. Currently, only the IFM-CAP model simulates single farms across the EU drawing on FADN. Although an important tool, IFM-CAP includes a rather coarse representation of alternative technologies, largely due to data limitations.
Promising tools, such as agent-based models (ABMs) so far face difficulties with application at larger scale. In the framework of the RUR-04-2018 published by the European Commission as part of the H2020 programme, three projects (AGRICORE - Grant agreement ID: 816078, BESTMAP - Grant agreement ID: 817501, MIND STEP - Grant agreement ID: 817566) received funding to address the challenges as stated above, namely to develop new data, tools and models that are needed to evaluate policy impacts on environmental and economic performance at the farm, regional, national and global levels. These new data, tools and models should start from the behaviour of the Individual Decision Making (IDM) unit in the agricultural sector and should include spatial and biophysical information and interactions between individual farms. The new IDM models or farm models should also be flexible and sustainable in use (keeping complexity within certain limits).
The first contribution discusses that this requires some degree of modularity where functionality can be added with additional models and data as needs arise.
The second contribution discusses calibration of ABMs to analyse adoption of ESS. Initial results from qualitative and quantitative interviews with farmers across five case study areas in the UK, Germany, Catalonia, Czech Republic and Serbia will be presented.
The third contribution focusses on harmonising and combining data from different sources, including data related to individual decision makers and how these can be used respecting privacy regulations.
The fourth contribution discusses the implementation and calibration of different types of behaviour in farm level models, following behavioural economic theories.
Authors: Wolfgang Britz (University Bonn), Alexander Gocht (vTI), Pavel Ciaian (JRC), Marc Mueller (WEcR)
The recent CAP reform allows and encourages the design of policy measures with more flexibility as was previously the case. This requires more detailed and flexible representation of policies in simulation models, taking the characteristics of the farm and its surroundings into account. Alternative specifications of behaviour and preferences of economic agents, and hence their willingness or ability to participate in policy programs, has received increased attention by policy makers and analysts alike. These requirements cause new challenges for the development of ex-ante simulation models of individual farmer’s decisions. Model development often takes place within the comparatively tight time- and budget lines of third-party projects, involving heterogeneous and changing research partners. This calls for a modular approach to model design, by which an existing core model is extended based on present project demands. Building on the experience made in previous projects related to integrated or modular model design, we review farm-level models currently in use for policy analyses in the EU. We focus on models that are developed or maintained by larger groups to investigate how collaboration is realized. We compare the findings with existing concepts for modular model designs form other domains and highlight some implications for future research.
Author: Guy Ziv, University of Leeds
Half of the European Union (EU) land and the livelihood of 10 million farmers is threatened by unsustainable land use intensification, land abandonment and climate change. Policy instruments, including the EU Common Agricultural Policy (CAP) have so far failed to stop this environmental degradation. The Horizon 2020 RUR-04 BESTMAP project will: 1) Develop a behavioural theoretical modelling framework to take into account complexity of farmers’ decision-making; 2) Develop, adapt and customize a suite of open source, flexible, interoperable and customisable computer models linked to existing data e.g. LPIS/IACS and remote sensing e.g. Sentinel-2; 3) Link economic, individual-farm agent based, biophysical ecosystem services and biodiversity and geostatistical socio-economic models; 4) Produce a simple-to-use dashboard to compare scenarios of Agri-Environmental Schemes adoption; 5) Improve the effectiveness of future EU rural policies’ design, monitoring and implementation. This talk will present the current conceptual framework for BESTMAP and initial results from qualitative and quantitative interviews with farmers across five case study areas in the UK, Germany, Catalonia, Czech Republic and Serbia.
Authors: Prof. Vangelis Tzouvelekas, Department of Economics, University of Crete, Greece; Dr. Pablo Baez, IDENER, Sevilla, Spain; Dr Michail Tsagris, Department of Economics, University of Crete, Greece; Prof Konstantinos Mattas, Department of Agricultural Economics, Aristotle University of Thessaloniki, email@example.com; Mr Carlos Leyva Guerrero, IDENER; Prof. Filippo Arfini, Department of Economics and Management, Università di Parma, Parma, Italy; Dr. Federico Antonioli, Department of Economics and Management, Università di Parma, Parma, Italy; Prof. Michele Donati, Department of Chemistry, Life Sciences and Environmental Sustainability, Università di Parma, Parma, Italy; Prof. Marco Riani, Department of Economics and Management, Università di Parma, Parma, Italy; Dr. Mario Veneziani, Department of Economics and Management, Università di Parma, Parma, Italy
Empirical analyses can take advantage of several data sets, covering different units of analysis, providing information regarding different characteristics, including those relating the different economic sectors in an area/region or nation. Datasets may have different temporal spans and data frequency as well as may be publicly or privately available, following the existence of privacy concerns (i.e., OECD vs FADN data).Researchers willing to take advantage of this wealth of data often struggle to harmonise, standardise and collate them from different sources, while maintaining the statistical representativeness of their characteristics and achieving the anonymous nature of the information. Researchers from the AGRICORE Project (Grant agreement ID: 816078) will present a data extraction and fusion module which is being developed to produce synthetic populations mimicking the distribution and characteristics of the real populations of interest, by means of a combinatorial optimisation method. Selected groups of agents -identified as being part of the synthetic population- are tested for their statistical match to pre-defined characteristics of the population of interest. The method is of interest to quantitative researchers who want to create a statistically representative population, arising from different datasets, and perform multiple/repeated quantitative analyses.
Authors: Scarlett Wang, Frederic Ang and Alfons Oude Lansink, Business Economics group, Wageningen University
An increasing amount of empirical evidence shows that farmers do not take profit maximization as their sole objective. Farmers may maintain the production at a sub-optimal level because they do not want to lose the non-pecuniary benefits. Clearly, costs and returns are important, but non-pecuniary benefits may make certain options more attractive than others (Howley, 2015). Typical farm level model like FARMDYN with maximizing net present value as objective, and assuming farmers’ rationality and risk neutrality may have limited prediction power for the real-life decision (Schaefer, Britz and Kuhn, 2017). This research aims to enrich the FARMDYN model with behavioural aspects, by estimating farmers’ risk attitude based on cumulative prospect theory, and explore the use of multi-objective programming.