=Paper=
{{Paper
|id=Vol-2030/HAICTA_2017_paper73
|storemode=property
|title=A Decision Support System for CAP2020 Regionalization Design in National Level
|pdfUrl=https://ceur-ws.org/Vol-2030/HAICTA_2017_paper73.pdf
|volume=Vol-2030
|authors=Dimitris Kremmydas,Michael Maliappis,Leyteris Nellas,Apostolos Polymeros,Stelios Rozakis,Kostantinos Tsiboukas
|dblpUrl=https://dblp.org/rec/conf/haicta/KremmydasMNPRT17
}}
==A Decision Support System for CAP2020 Regionalization Design in National Level==
https://ceur-ws.org/Vol-2030/HAICTA_2017_paper73.pdf
A decision support system for CAP2020 regionalization
design in national level
Dimitris Kremmydas1,†, Michael Malliapis1, Leyteris Nellas1, Apostolos Polymeros3,
Stelios Rozakis2, Kostas Tsiboukas1
1
Department of Agricultural Economics & Rural Development, Agricultural University of
Athens, Ieara Odos 75 118 55, {kremmydas,michael,enellas,tsiboukas}@aua.gr
2
Dept.of Bioeconomy and Systems Analysis, Institute of Soil Science and Plant Cultivation,
Poland, srozakis@isc.tuc.gr
3
Directorate of Agricultural Extension, Greek Ministry of Agriculture,
apolymeros@minagric.gr
Abstract. The latest Common Agricultural Policy reform provides national
authorities with several implementation options for fine tuning individual
goals. Among other, member states can opt for regionalization, i.e. vary the
basic payment unit value between national agronomic or administrative regions
that have been defined at the beginning of the programming period. We
propose a Decision support System that will support national authorities to
implement regionalization in a transparent way facilitating collaboration with
different shareholders. In this paper we present an overview and give a proof-
of-concept implementation.
Keywords: Common Agricultural Policy, Decision Support System; Basic
Payment Scheme
1 Introduction
The latest Common Agricultural Policy reform (CAP2020) provides national
authorities with several implementation options for fine tuning individual goals. 30%
of the national CAP funding is connected to the farmers’ compliance to a predefined
set of pro-environmental practices. Up to 5% can be devoted to farms of areas with
natural constraints, up to 13% to couple payments, up to 10% to small farmers’
scheme, up to 2% to new farmers’ scheme and up to 3% to the national rights stock.
The rest, called basic payment scheme (BPS) is the main layer of income support
(over 50% of the national budget), based on payment entitlements activated on
eligible land and decoupled from production.
Within this scheme, among other options, Member States (MS) can opt for
regionalization, i.e. vary the BPS unit value between national agronomic or
administrative regions that have been defined at the beginning of the programming
period. Indeed, six MS (Germany, Greece, Spain, France, Finland and United
Kingdom) have enabled this option while the rest have applied a uniform national
BPS unit value (Henke et al., 2015). Among the regionalization countries only
602
Greece has used a purely agronomic criterion while the rest used administrative
regions.
The only regulation guideline regarding regionalization is that it should be in
accordance with objective and non-discriminatory criteria, such as their agronomic
and socio-economic characteristics, their regional agricultural potential, or their
institutional or administrative structure (European Parliament, 2013). Practically MS
are totally free to draw the regions and allocate the BPS budget.
This flexibility provides to policy makers numerous different alternatives on how
to form regions and allocate the budget. Additionally the fact that different
stakeholders are affected in a distinct way, call for a transparent design process.
Towards this end we propose a Decision Support System (DSS) that will support
national authorities to implement regionalization in a transparent way facilitating
collaboration with different shareholders. In this paper we present an overview and
give a proof-of-concept implementation.
2 CAP2020 Regionalization
Regionalization is composed of two distinct components:
Drawing the different regions: As already noted, the distinction of regions can be
based on administrative, agronomic or both dimensions.
In the first case, the basic payment of the farm in euro/ha (BP!! ) equals to the basic
!
payment of the administrative region that the farms belongs to (𝐵𝑃!(!) ), as in
equation 1.
!
𝐵𝑃!! = 𝐵𝑃!(!) (1)
In agronomic-based regionalization this equals to the sum of the basic payment for
each crop family (agronomic) weighted by their relative area share to the total
utilized agricultural area of the farm in a baseline year, as in equation 2. In this
second case farms do not form clear partitions, i.e. one farm may be affected by more
than one agronomic region.
!
! !(!) !"! ∙!!,! (2)
BP!! = !
!"!
Allocating the budget between regions: A national basic payment budget (B) has
to be divided among regions. If 𝐴! is the percentage of budget allocated to region r,
then the basic payment unit value equals to the total amount allocated to the region
(B ∙ A ! ) divided by the total eligible land of the region as in equation 3.
!
BP!! = B ∙ A ! !(!) TL!(!) (3)
603
3 An Overview of the Decision Support System
National authorities have a great flexibility on how to draw regions and allocate
budget. Consequently they can potentially pursue a wide range of objectives. This
means that the required data in order to evaluate the objectives can only
approximately be determined during the development of the DSS and very probably
new data will be requested during the consultation with other stakeholders.
The DSS shall address this data flexibility requirement by using a Data
Warehouse (DW) solution, which is ideal for gathering and keeping policy data from
many independent sources of information (Malliapis & Kremmydas, 2016). The
underlying DW will contain socio-economic data with spatial and administrative
dimensions attached. For instance the database could contain the current direct
payment allocation per farm size and prefecture from the National Payments
authority; the income indicators per farm size and farm activity from the national
Farm Accounting Database Network microeconomic database (FADN); the regional
GDP per sector from the national statistical authorities; etc.
Nevertheless, given an established strategic goal, the DSS shall provide a clear
picture of how that goal is affected for any selected scenarios. The DSS usage is
expected to be in an iterative mode: policy makers and stakeholders draw regions, try
some budget allocations and observe the effects and then restart the process to fine
tune policy results.
Thus we distinguish three DSS use cases that correspond to the above workflow:
Exploratory region formation use case: Policy makers form a regionalization
scenario, i.e. define regions, by means of exploratory analysis. The definition of
regions is based on some partition variables, e.g. the NUTS nomenclature, the
altitude or some crop classification like arable vs. permanent crops. Thus the user
selects the partition variables, which identify the different regions. The user very
probably will further consolidate those regions to more homogeneous ones. In order
to do so, he will examine certain regions’ property variables, e.g. the prevailed crop
pattern, the importance of agriculture, the current single payment unit value, etc. He
can thus refine initial region creation either manually or through a clustering tool that
will suggest him the regions that are as homogeneous as possible. The activity
diagram of this use case is provided in Figure 1.
Budget allocation effects use case: When the user has concluded to a region
formation scenario he is ready to set budget allocation. This is expected to be a trial
and error exercise where policy makers investigate the effects across different
stakeholders for different allocations. Users can either manually set the budget share
or can use tools of predefined allocations, e.g. budget share proportional to the
number of entitlements or to the gross value of direct payments in each region. Then
the DSS engine will calculate the indicators and present them to the user. Based on
the results the user can save the regionalization scenario and restart the process. The
indicators of the scenario effect will span to different stakeholder classes, e.g. farms
per NUTS administrative level or per type of farming or per farm income class. The
activity diagram of this use case is provided in Figure 2.
604
Dissemination and collaboration use case: Due to the different interests of
stakeholders, collaboration is a very important aspect of the regionalization decision
process and thus is incorporated into the DSS. When a user is satisfied with a
scenario (region definition + budget allocation) he can save it and choose to share it,
either with other users of the system or in public. A discussion channel, e.g. a forum
thread, will be automatically created so that other users can comment and discuss
scenarios. Users will also be able to load the scenarios of other users in order to
adjust them to their point of view.
Fig. 1. Activity Diagram for exploratory extraction of regions
605
Fig. 2. Activity Diagram for exploring the effects of various budget allocations between
regions
4 A proof-of-concept implementation
We used the R-Shiny web application framework (Chang et al., 2017) for agile
development of a proof-of-concept implementation. The data source was the Greek
FADN database without resorting to a Data Warehouse solution.
For the exploratory region formation stage, we used the following partition
variables: NUTS-2, NUTS-3, Altitude and the Less-Favored-Area status. For
deciding on the region homogeneity we provided the following property variables:
606
Farm Gross Income, Farm Gross Income per ha, cereals and permanent crops area,
percentage of subsidies on gross income, Single Farm Payment value, sum of Single
Farm Payments, Irrigated area. A screenshot of this first use case can be seen in
Figure 3. User selects partition and property variables and by clicking GET STATUS
he gets an information table where he can further refine region formation. There is
also a CLUSTER VARIABLES button that triggers a hierarchical clustering of the
regions.
Fig. 3. Exploratory region formation use case
When he is satisfied with the selected region formation scenario he can click
SELECT PARTITION that will take him to the next step (Figure 4). There he can set
the national budget and adjust the allocation to regions. Then by pressing UPDATE
ALLOCATION, the DSS calculates the effects and present him results in various
ways. Administrative level gross income positive or negative effects are shown in a
NUTS-3 map. A table of the effects on different type-of-farming within a NUTS-2
context and on different farm size classes is also given in a different tab.
607
Fig. 4. Budget allocation use case
5 Conclusions
In this paper we present an overview of a Decision Support System for assisting in
the design of EU-CAP regionalization scheme. The proof-of-concept implementation
shows that the DSS can greatly enhance the policy design process and provide
insights to policy makers. Furthermore it can serve the cause of transparency to
different shareholders.
The next step is to implement the DSS in a real case study, where in collaboration
with the policy makers and the stakeholders, different regionalization scenarios and
budget allocations will be drawn, the various trade-offs will be revealed and a
consensus will be reached, employing group decision making methods.
References
1. Henke, R., D’Andrea, M. P., & Benos, T. (2015). Implementation of the first
pillar of the CAP 2014 – 2020 in the EU Member States,
http://doi.org/10.2861/662407
608
2. European Parliament, C. of the E. U. (2013). EU 1307/2013: Regulation (EU) No
1307/2013 of the European Parliament and of the Council of 17 december 2013
establishing rules for direct payments to farmers under support schemes within
the framework of the common agricultural policy and repealing Counc. Official
Journal of the European Union, 2008(347), 608–670.
https://doi.org/10.1017/CBO9781107415324.004
3. Maliappis, Michael T & Kremmydas, Dimitris (2016). Data warehouse
technology for agricultural policy data: a Greek case study. International Journal
of Sustainable Agricultural Management and Informatics, 2, 243-262.
4. Chang, W., Cheng, J., Allaire, J. J., Xie, Y., & McPherson, J. (2017). shiny: Web
Application Framework for R. Retrieved from https://cran.r-
project.org/package=shin
609