The objective of this case study is to provide practical examples from the European context of how digital technologies can improve systems which pay farmers for producing ecosystem services.
Digital Opportunities for Better Agricultural Policies
Chapter 9. Case Study 4. Earth Observation initiatives for administration of the European Union Common Agricultural Policy
Abstract
Context: Reforming the CAP’s administration
The European Union (EU) Common Agricultural Policy (CAP) is the overarching system of subsidies and payment programmes for agriculture and rural areas in the European Union. The CAP pursues a range of objectives,1 and accounts for over 40% of the European Union’s annual budget.
Schmedtmann and Campagnolo (2015, p. 9326[1]) provide an overview of the administrative mechanisms of the CAP:
To ensure that CAP funds are spent appropriately, Member State Authorities have to comply with legal management and control mechanisms…Each Member State is responsible for subsidy administration and control, which are carried out by a National Control and Paying Agency (NCPA).
In order to obtain area-based financial support [direct payments], farmers are required to submit an application to their NCPA early in the year, where they declare the precise location of all of their agricultural parcels, as well as the crop type. The National Agency is responsible for controlling at least 5% of those declarations and penalizing farmers who submit incorrect information by performing so-called On-The-Spot (OTS) checks. For area-based subsidies, an agricultural parcel must be controlled at two different levels: both the declared crop and area must be correct. The [European Commission] in turn controls the NCPAs. When discrepancies between the control result and the reality are found, a Member State is penalised and has to return to the EU part of the subsidies that were distributed to farmers.
The complex process of subsidy control requires computational tools: NCPAs rely on Integrated Administration and Control System (IACS), which includes a [Land Parcel Identification System] LPIS. The main functions of those spatial databases are localization, identification and quantification of agricultural land via detailed geospatial data, in order to facilitate the distribution of CAP subsidies.
Following the 2013 CAP reform, EU farmers are able to apply for direct payments through the Basic Payments Scheme (BPS) (OECD, 2017[2]). These payments are intended to act as a safety net in the form of a basic income support. Cross compliance and Greening are two mechanisms (referring to specific obligations) that are linked to this payment to ensure more environmentally-friendly farming approaches and deliver continued food security and safety in Europe. The introduction of cross compliance and greening measures introduces additional complexity for programme administrators.
Use of digital technologies to streamline CAP administration
The problems
The CAP is fundamentally an eligibility-based system: while the conditions have changed over the years (particularly with the decoupling reforms in the mid-2000s), farmers must still meet certain criteria in order to receive payments. As in regulatory contexts more generally, the eligibility system introduces a potential incentive misalignment problem: while the administrator has incentive to discover whether the farmer meets the criteria, absent the threat of sanctions,2 a farmer would prefer to receive the payment without incurring costs needed to achieve eligibility for the payment. This creates incentives for farmers to preserve information asymmetries between themselves and the administrator (i.e. the farmer knows his or her own action but may have incentive to prevent the administrator from accessing that information). The reform of the CAP to include environmental greening and cross compliance requirements, which may be costly for farmers to meet, exacerbates this potential. Therefore, a system of monitoring, controls and sanctions is needed to ensure the effectiveness of the payment as an incentive mechanism for improving the sustainability of European agriculture.
As outlined above, there are three main administration and control tools used by the relevant competent public authorities (“National Control and Paying Agencies”, NCPA): administrative checks of paperwork submitted by claimants (farmers), physical on-the-spot checks (OTSC) and Control with Remote Sensing (CwRS). Due to the high complexity and diversity of the obligations that need to be verified, each method has its limitations. As a result, existing administration and control regimes entail high transaction costs for public administrators as well as private transaction costs and administrative burden for farmers. For example, according to DG-AGRI (Borchmann, n.d.[3]), the cost of controls to Member States in 2015 was EUR 1 125 million, which equates to 5.2% of total public CAP expenditure.
The challenge for CAP administrators is therefore to minimise administrative transactions costs (both public and private) while maintaining effective standards of compliance.3 One crucial aspect of this is to reduce costs of obtaining information on farmers’ activities.
Digital solutions: The RECAP initiative
The use of digital technologies, particularly earth observation technologies and online digital platforms, offers the potential to provide improved monitoring of agricultural activities at lower cost than existing administration and control methods. While there are numerous initiatives research and testing of digital solutions aimed at reducing the costs of administering the CAP while increasing information on farmers’ activities,4 this case study centres on one initiative from the European context: the RECAP—Personalised Public Services in Support for the implementation of the Common Agriculture Policy (CAP), an Horizon 2020 project funded under the ICT-enabled open government (H2020-INSO-2015-CNECT) call (Grant Agreement 693171), which aims to provide practical evidence on these potential benefits.
The initiative commenced in May 2016 with 30 months duration, involving 12 partners.5 The overall budget of the project is EUR 2.7 million (EUR 2.1 million requested EU contribution). It is based on the following interrelated objectives:
To develop improved e-public services that enable a better implementation of the CAP and simplify administrative procedures, integrating open and user-generated data.
To develop personalised public services that support farmers to better comply with CAP requirements.
To increase the transparency of compliance monitoring procedures related to CAP.
To enable the reuse of data (open and user-generated) by agricultural consultants and developers for delivering their own added value services for farmers.6
To pilot test the services in an operational environment with the participation of end users in five countries (Greece, Lithuania, Serbia, Spain and the United Kingdom).
To assess the usability, effectiveness and impact of the proposed services in delivering the public administrations’ goals, and provide feedback into a set of recommendations for future use of these approaches to deliver more effective and applied public administration.
RECAP is a commercial platform7 (cloud-based Software as a Service (SaaS)) that integrates satellite remote sensing and user-generated data into added value services for public authorities (administrators and inspectors), farmers and agricultural consultants, to improve the processes for implementing and monitoring the BPS. RECAP has three interrelated results indicators:
Increasing the efficiency and transparency of public authorities’ procedures implementing and monitoring the CAP by enabling effective and efficient remote monitoring of farmers’ obligations (including automation of compliance checks for some requirements) through the use of open earth observation (EO), user-generated data (geotagged and timestamped photos) and purpose-built algorithms. The RECAP pilots aim to reduce administrator costs by at least 25% (Table 9.2).
Providing personalised services to farmers for their better compliance with the CAP environmental standards (Cross Compliance (CC) and Greening Measures). The RECAP pilots aim to reduce farmer administration costs related to BPS claims by at least 25% (Table 9.2).
Stimulating the development of new added value services by agricultural consultants and developers who can create add-ons to the main platform and make use of the data collected.
RECAP digital components
To achieve these deliverables, RECAP makes use of various digital tools or “components”, explained below (see also Table 9.1).
Remote Sensing component
The RECAP Remote Sensing (RS) component provides automated earth observation processing workflows to assist paying agency inspections of farmers’ compliance with their CAP obligations. The methodology is founded on the accurate crop type classification via applying a machine learning algorithm to a time-series of combined Sentinel‑2 imagery and relevant vegetation indices. The monitoring of compliance was algorithmically addressed for specific Cross-Compliance and Greening requirements8 (Box 9.1).
The practicality of the output RS information ranges from direct decision making (e.g. for crop diversification) to simple indicators of potential noncompliance (e.g. for minimising soil erosion), depending on the complexity of the individual CAP obligation. The RS component comprises three principal processing chains:
crop type mapping (classification)
runoff risk analysis
identification of stubble burning.
Figure 9.1. RECAP initiative components
Source: RECAP initiative case study participants.
The relevance of the developed RS solution to the CAP monitoring challenge is essentially based on the accuracy of the crop type classification. Accuracy was assessed by comparing classification based on satellite data obtained in two iterations9 against validated ground truth data obtained by pilot inspections in selected subsets of the dataset.10 As an example, the accuracy results obtained in the Spanish pilot are presented in Table 9.1.Validated results showed an overall crop type mapping accuracy in the range 80-90% for the identification of 9-13 different crop types, depending on the pilot, which explain more than 90% of the regional agricultural zone (Source: Personal communication, case study participants, July 2018).
Table 9.1. Accuracy of RECAP Remote Sensing component crop classification, Spanish pilot
1. Producer's Accuracy is the map accuracy from the point of view of the mapmaker (the producer). This is how often are real features on the ground correctly shown on the classified map or the probability that a certain land cover of an area on the ground is classified as such.
2. The User's Accuracy is the accuracy from the point of view of a map user, not the mapmaker. The User's accuracy essentially reveals how often the class on the map will actually be present on the ground.
Source: Draxis Environmental (2019[4]).
Overall, the crop classification algorithm was assessed to provide satisfactory results: 75-85% accuracy, even for datasets that include satellite imagery only until mid-late June. This is very important, since paying agencies require accurate information at the time of the farmers’ applications, in order to better target their sampled on-the-spot inspections to parcels that constitute potential breaches of compliance.
Crop classification from the RS component was provided also with the crop type as declared by the farmers. This functionality is key for paying agencies, as (together with the ground-truthing accuracy); it allows probabilistic identification of potential non-compliance. The RECAP team developed a “traffic light system” to convey this probabilistic assessment in an intuitive way.11 Where the ground-truthing accuracy of the RS classification is high, but the RS classification disagreed with the declared classification, this indicates potential non-compliance (untruthful declaration).Towards the completion of the project, validated results were received for each of the three pilots as follows:
Spain: Of 107 random parcels inspected, 105 were classified correctly.
Greece: Inspectors visited only parcels that were selected by the smart sampling methodology; that is, parcels classified with high confidence to crop types other than the one declared, which are considered as potential breaches of compliance. It was shown that 76 out of 85 inspected parcels were indeed wrongly declared and correctly classified.
Lithuania: The validated dataset acquired through the Lithuanian Paying Agency inspections revealed an actual overall accuracy of 76.2% in late June and 80% in late August out of 3 319 parcels inspected.
The crop type classification accuracy, and hence the usefulness of the RS component for CAP administrative decisions for which crop-type is relevant, depends on three parameters:
Percentage of truthful declarations
Cloud coverage
Parcel size
In one of the case studies—Navarra, Spain—where 90% thematic accuracy was achieved, all these parameters were optimal. This means more than 97% of truthful declarations, limited cloud coverage, and an average parcel size of 2 ha, which is considered sufficiently large for a Sentinel-2 based classification.
When a considerable percentage of declarations are not truthful, then similar crop types, both in spectral characteristics and phenology—e.g. wheat, barley, oats—might not be well discriminated. Hence, merging of such crop types into spectral coherent clusters (e.g. cereals) would be necessary for an adequately accurate result. Therefore, the thematic accuracy of the crop identification products depends on the type of information one is aiming for. For example, the usefulness of the clusters in assessing a crop rotation requirement depends on the degree to which farmers could implement a crop rotation within, as opposed to across, clusters.12
Crop classification accuracy also depends on the size and shape of the parcel, with classifications for larger parcels and parcels with straighter borders tending to have higher accuracy than smaller parcels or parcels with more irregularly-shaped boundaries. The parcel area is important since accuracy depends on the number of image pixels that fall within the parcel boundaries. Sentinel-2’s 10 m pixel size equates to 50 image pixels in 0.5 ha of land. An analysis conducted, comparing the accuracy of classification in conjunction with the parcels’ size, showed that having 50 pixels of information provides accurate results, whereas for smaller parcels the decision is both less confident and less accurate.
Box 9.1. Evaluation of remote sensing (RS) and machine learning (ML) tools to classify crop types and monitor environmental requirements
RECAP case study participants commented on the practicalities of using RS information and machine learning to successfully classify crop types and identify compliance with environmental requirements (e.g. GAEC, greening):
“Different description of crop types would imply different spectral signatures for the crop classes and thereby different classification results. Additionally there are differences in the percentage of correctly declared cultivated crop types that accordingly affect the training of the machine learning algorithms. In Navarra, Spain declarations are almost completely correct and therefore results are excellent. In Greece, however, there is a significant percentage of falsely declared crop types that affects the classification accuracy. Nonetheless, the algorithm is indeed robust; in the sense that if 20% of declarations are wrongly stated this would roughly mean only 5% reduction in accuracy. Finally, in countries such as Lithuania, where cloud coverage is significant throughout the year algorithmic modifications are necessary. For example, it was found that a different machine learning algorithm performed best for the Lithuanian case.
The main pillar of the agriculture monitoring scheme is the accurate crop type classification. The practicality of the classification is straightforward. However, RECAP attempted to specifically address the compliance of farmers to their actual CAP obligations (GAECs, SMRs, Greening). For some CAP obligations, such as Greening 1, crop classification is indeed all that is needed to decide on the compliance of the farmers. Now, for other obligations such SMR 1 (Reduce water pollution in nitrate vulnerable zones), the RS component of the RECAP platform provides a risk assessment on the soil loss and runoff to nearby watercourses, for each parcel. This is indeed a prerequisite for the farmers in order to comply with SMR 1, but the rule also extends to manure spreading obligations that cannot be addressed by remote sensing. Therefore, even though the remote sensing information provided with respect to SMR 1 is useful, it is not complete for compliance decision making.”
Source: Case study participant, Dimitrios Petalios (CREVIS) (June 2018).
Spatial component
The spatial component depicts in digital maps (set in several layers) the information generated by the RS component as well as external spatial data, which are listed below. These maps enable users to visualise valuable information for an effective and efficient inspection process (Paying Agencies (PAs), inspectors) (Table 9.2 provides an example of how the PAs view the spatial component). The produced maps include:
Time-series of Sentinel-2 true color composites and vegetation indices (viewing only)
Natural habitat sites
Nitrate Vulnerable Zones
Botanical Heritage Sites
Watercourse maps
Slope map
Administrative boundaries and settlements
Land Use/ Land Cover Maps
Land Parcel Identification System (LPIS)
RS-derived parcel specific thematic information (i.e. crop type mapping, polluted water runoff risk assessment, identification of stubble burning, soil erosion, etc.). This is displayed in the form of a list when interactively selecting the parcel of interest.
Figure 9.2. RECAP spatial component
Source: RECAP initiative of case study participants.
The remote sensing results and the information provided by the Spatial component can be used by the PA as auxiliary information in their risk analysis process and identify the farmers that are more likely not to comply, so that they could proceed with more targeted inspections. Specifically, the PAs are able to draw a bounding box on the map, covering the area of interest, and for which they will receive the remote sensing analysis results.The crop classification information is also provided to the farmers through their own profile. Therefore, if their parcel is classified to a different crop type than the one declared and with high confidence, then they could opt to change their declarations if they agree with the classification.
Business intelligence component
The Business intelligence component is aimed at the PA officers only. It is a data-mining tool enabling public officers to analyse large datasets stored in RECAP platform. PAs will be able to make use of available data and create key performance indicators (i.e. CAP objectives) and relevant reports, enabling them to set targets and move towards a more result-oriented CAP support. Additionally, this tool allows PAs to extract valuable information from such vast datasets and to uncover previously unknown patterns that may be relevant to current agricultural problems, thereby helping farmers and managing organizations to transform data into business decisions and ensure a better implementation of the CAP.
Workflow component
The workflow component is the core system of the platform, working as the link between the different parts of the system. Specifically, it brings together information that is processed by all components to the user in a way that is easy to understand. It functions as an orchestrator for the RECAP business logic, the communication with the data storage, the Application Programming Interface (API), the receipt of information from outside sources and the validation process. For example, it provides farmers, consultants and inspectors with checklists of Cross Compliance rules applicable to the farm, based on information from the BPS application submitted by the farmer; it guides farmers and inspectors with personalised information on the procedures to follow regarding the compliance procedure; generates notifications to farmers based on calendar of key dates.
Software Development Kit
The Software Development Kit (SDK) allows agricultural consultants and developers to develop their own “added value” services in an open approach within the RECAP platform, and deliver improved advisory services to farmers. The SDK enhances the role of the platform, both by enabling consultants to develop their own services on top of the RECAP platform using design tools, libraries and communications with the database under an open approach; and also by supporting any technical integration with external systems.
The RECAP Digital Platform—Web and mobile application
The RECAP platform interconnects the different system components in order to deliver the deliverables earlier described. Being co-created and co-produced with its end-users, it covers five categories; the general system requirements, the Basic Payment Scheme (BPS) eligibility and applications, the farmer record keeping, the inspection process and the farmer education and information. The main features covered per category (table rows) and user group (table column) are presented in Table 9.2.
Table 9.2. Main features of the RECAP Platform (web application), by user group
Note: SDK = Software Development Kit.
Source: RECAP initiative case study participants.
Apart from the web-based application platform, two mobile interfaces are developed; a smartphone-optimised interface dedicated to the farmers’ needs and another one focusing on the inspectors’ needs. The mobile application is mainly for the data collection on the farmer’s field either from the farmer or from the inspector during on-the-spot checks (to overcome connectivity challenges, the mobile application also has an offline mode – when operating in offline mode, data will be uploaded to the RECAP database once the mobile application is re-connected to internet).
On the RECAP platform, each farmer has their own personal account where they are able to store data, records, and documents that need to be obtained or retained. This can be presented to inspectors during an inspection. The RECAP platform allows the farmer to filter complex cross compliance rules and see only those relevant to their farm. There are also alerts for actions to be taken and potential non-conformities. These alerts provide farmers access to checklists and workflows through a mobile and web interface (the RECAP platform). The PA is responsible for updating and certifying the checklist(s).
Satellite imagery is available for all users of the platform. However, PAs are able to see the “big picture” (all parcels within the user-defined area of interest), while farmers and consultants see cropped imagery that includes only the farmers’ parcels.
The digital platform also enables PAs to increase the effectiveness of risk-based analysis for the selection of farms to be inspected through the help of the Remote Sensing component, which uses a combination of open and user-generated data. The RECAP platform allows the PAs to select a farm for inspection and retrieve farm profile data and previous inspection results, which will be available both to the PAs and to inspectors. Overall, RECAP delivers a platform to public administrations so that they carry out inspections more efficiently, more accurately and more quickly.
Implementing the RECAP pilot
The RECAP platform is currently being tested and validated in an operational environment in five countries—Greece, Lithuania, Serbia, Spain and the United Kingdom—with the active participation of public organisations, agricultural consultants, and farmers. The platform is comprised of five different workflows (one for each country pilot), due to the differences between the pilots and the CAP rules interpretation.13 Based on this, the RECAP platform is developed as an integrated system, composed of core functionalities that are commonly shared across the pilots, with additional pilot-specific functionalities are built on top of these core functionalities.
Pilot implementation in Spain, Greece, and Lithuania is focusing on delivery of public services, with the participation of four public organisations (Paying Agencies and Agricultural Advisory Services) which are members of the project consortium (INTIA, OPEKEPE, NMA, and LAAS). In the United Kingdom, the pilot implementation focuses on delivery of personalised services from agricultural consultants (partners STRUTT & PARKER).
The Serbian Pilot (INO) case is focused on organic agriculture, with organic certification bodies, organic farmers and public bodies to overlook that organic certification is in line with legal requirements (Official Gazette RS 30/2010; fully aligned with EU regulation on organic farming – Regulation EC 834/2007). The RECAP platform will support the entire process of subsidy provision for organic farmers, certification agencies, agricultural consultants and for the public authorities tasked with implementing, managing and controlling this payment scheme. Serbia being an EU candidate country (2012), has started accession negotiations in 2014 and is committed to transpose and implement the acquis14 on agriculture and rural development by the date of accession. RECAP platform will be positioned to support monitoring aspects of relevant subsidies within the Instrument Pre-Accession Assistance in Rural Development (IPARD) and providing assistance for the implementation of the acquis concerning the CAP.
The outcomes of the Pilot contribute to the achievement of the strategic impacts of the RECAP project: the stimulation of the creation, delivery and use of new services coupled with open public data; the delivery of more personalised public services that better suit the needs of users; the reduction of the administrative burden of citizens and businesses and the increased transparency of and trust in public administrations. The achievement of the impacts was measured through the monitoring of a set of result and impact indicators (Table 9.3).
Table 9.3. Result and impact indicators for the Pilots
Source: RECAP initiative case study participants.
A monitoring and evaluation system (qualitative and quantitative tools) was used to ensure on the one hand the proper development of the Pilots to achieve the expected outcomes, and, on the other hand, to allow assessment of whether the specified result and impact indicators have been achieved and obtain relevant inputs for the RECAP solution sustainability and future adaptations.
The first three targets (number of farms participating in pilots, number of cross-compliance inspections carried out using RECAP, number of OTSC carried out using RECAP) have all been achieved. Upon completion of the pilot, participants were surveyed about their perceptions about the extent to which the RECAP platform reduces administrative burden and facilitates compliance. Selected results from this survey (RECAP Consortium, 2018, p. section 4.2[5]) are:
61% of farmers participating in the RECAP pilot somewhat agreed or strongly agreed that the RECAP platform increases their understanding of CAP Cross-Compliance (CC) rules, and 55% somewhat or strongly agreed that the platform decreases the likelihood of their breaking CC rules.
42% of agricultural consultants participating in the pilot reported that the necessary time for preparing Basic Payment Scheme (BPS) application will be shorter using the platform; and the corresponding time reduction is >25% for 60% of this subset; the remaining 44% considered that time spent preparing applications would not change. Similar results were found in relation to time spent checking adherence to CC rules.
51% of participating farmers considered that their necessary time for preparing a Basic Payment Scheme (BPS) application would be shorter using the platform (and 64% of this subset of farmers considered that the corresponding time reduction would be greater than 25%); compared to 44% of farmers who considered the time spent would not change, and 5% who considered their time spent making an application would be longer. Similar results were found in relation to time spent checking adherence to CC rules.
82% of organic farmers somewhat or strongly agree that the platform increases their understanding of compliance with Organic Certification and Organic Subsidies; 77% believe it will help them to follow organic certification requirements, and 91% believe that using the system will reduce time for presenting evidence of compliance with Organic Certification requirements.
74% of inspectors somewhat or strongly agree that the platform makes the CC procedure more transparent, while 68% believe the platform increases the accuracy of OTSC for CC;
62% of inspectors consider the time spent inspecting a farmer would be shorter using RECAP, and of these, 60% considered the time reduction would be greater than 25%. Similar results were found in relation to the number of plots inspected per day.
58% of inspectors somewhat or strongly agree that the platform allows for the reduction of administrative burden for inspectors.
100% of certification bodies somewhat or strongly agree that the platform will assist them with Organic Certification and that it reduces administrative burden.
Future plans for RECAP and beyond
The outcomes of the RECAP pilots were presented at the European Unon’s 2018 INSPIRE Conference.15
While the RECAP initiative formally concluded in 2018, there are a number of further EU initiatives which, like RECAP, aim to simplify and modernise administration of the CAP.
Box 9.2. Further EU collaborative initiatives for innovative tools for CAP 2020+
The RECAP initiative, which commenced in 2016, is a forerunner in what has become a very active research and innovation space within the European Union. Since that time, a number of collaborative initiatives have commenced, which aim to encourage new tools and processes for modernising and simplifying the CAP in the next programming period (beginning 2020) and beyond. Key initiatives are:
Pilot4CAP is a platform for sharing Pilot projects for the new CAP2020+, hosted and coordinated in the G4CAP Web application. This platform calls for sharing and reporting of publicly known new or ongoing pilot projects performed in preparation for the new CAP 2020+. Projects on the following subjects can be entered: IACS, OTSC, LPIS, Land Use, Land Cover, (IT or other) services making use of imagery such as Sentinel optical, Sentinel radar, VHR/HR satellite, aerial photo, RPAS or High Altitude drones (HADs).
The Sentinels for Common Agricultural Policy (Sen4CAP) project, launched in May 2017, aims at “providing to the European and national stakeholders of the CAP validated algorithms, products, workflows and best practices for agriculture monitoring relevant for the management of the CAP. The project will pay particular attention to provide evidence how Sentinel derived information can support the modernization and simplification of the CAP in the post 2020 timeframe. Sen4CAP has been set up by ESA in direct collaboration and on request from DG-Agri, DG-Grow and DG-JRC.”
Lessons learned
Lesson 1. Earth-observation tools powering accessible, user-specific platforms offer the opportunity to substantially reduce transactions costs of administering the Common Agricultural Policy
As the results from the end-of-pilot survey showed, pilot participants (farmers, agricultural consultants, inspectors, certification bodies and national paying agencies) all generally considered that the RECAP platform would reduce administrative burden. In some cases, reductions in administrative costs (generally measured as time spent on various administrative activities) were considered to be greater than 25%.
Lesson 2. By using spatially-explicit earth observation and other data on a wide range of agricultural and environmental variables, RECAP paves the way for more nuanced, targeted agri-environmental policies
Beyond lowering the administrative costs of implementing existing CAP programmes and requirements, RECAP-style digital platforms based on earth observation data enables public authorities to better monitor the implementation of agricultural and agri-environmental policies, and paves the way for more targeted policies in the future. In particular, the provision and availability of highly-differentiated spatial data (e.g. by parcel) on agricultural practices and landscape characteristics (e.g. slope, proximity to receiving waters, soil type, etc.) at high temporal frequencies will allow agencies to pursue more spatially and dynamically flexible policies that were previously infeasible due to data constraints.
Lesson 3. Digital tools such as the RECAP platform can increase the transparency of inspections and the accountability of public organisations, resulting in greater robustness of, and trust in, public agencies
The RECAP platform provides access to frequently updated satellite data and to functions for inspectors or farmers so that they may upload geotagged, time-stamped images to support administrative checks of eligibility and compliance. Thereby, farmers have continuous access to further farm-related details within a secure and transparent framework. Further, farmers can use the images uploaded in a number of ways: e.g. share them with advisors and seek assistance or rectify non-compliance or prevent such a case occurring in the future.16 RECAP is therefore a tool that assists fair, transparent and detailed inspections.
Lesson 4. RECAP uses a co-operative approach to ensure the efficiency and effectiveness of its technical solutions, and interoperability with other solutions
RECAP helps to foster a less adversarial administrative context by “building bridges” between public administrators and farmers through the use of innovative Earth Observation solutions and related tools. It is based on a user-driven approach with its solutions having been designed and developed alongside the end-users and stakeholders, under a co-creation and co-production scheme.
The collaborative approach also encourages proactive participation of farmers in the overall monitoring procedure, giving them an active role in the data collection process, enhancing close communication and co-operation with public administration. This innovative approach sets up a monitoring system that informs, guides and notifies farmers on their obligations towards the BPS regulations, instead of penalising them for non-compliance when inspections take place.
Finally, RECAP also offers an Application Programming Interface (API) allowing to other platforms to use the RECAP data or contribute data to the RECAP database. This allows for interoperability and interconnectivity with other platforms or applications offered by PAs as well as ensures further integration with other systems developed (or to be developed) by agricultural consultants. In this way, RECAP allows for the “only once” principle, according to which information submitted once by the farmers need not be asked for again by another service of the administration.17
Lesson 5. Innovative digital solutions such as RECAP can underpin new private sector business models
The innovative solutions that the RECAP platform provides to agricultural consultants give rise to new business opportunities. Provided with the Software Development Kit, agricultural consultants are offered certain functionalities allowing them to search and use data stored in the RECAP platform; to integrate search results into their applications supporting farmers’ claims; and to manage RECAP configuration and objects. Overall, RECAP can be used as a tool to underpin the day-to-day work of agricultural consultants to provide valuable advice to farmers.
References
[3] Borchmann, C. (n.d.), CAP audit system 2014-2020, European Commission, DG-AGRI, https://ec.europa.eu/futurium/en/system/files/ged/c._borchmann_dg_agri_eafrd_audits_-_facts_and_figures.pdf (accessed on 20 August 2018).
[4] Draxis Environmental (2019), D4.4 RECAP Final Evaluation Report, https://www.recap-h2020.eu/pilots/ (accessed on 19 March 2019).
[6] European Commission (n.d.), The common agricultural policy at a glance, https://ec.europa.eu/info/food-farming-fisheries/key-policies/common-agricultural-policy/cap-glance_en (accessed on 20 August 2018).
[2] OECD (2017), Evaluation of Agricultural Policy Reforms in the European Union: The Common Agricultural Policy 2014-20, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264278783-en.
[5] RECAP Consortium (2018), Final evaluation report, http://ttps://www.recap-h2020.eu/wp-content/uploads/2018/12/d4.4._final_evaluation_report.pdf (accessed on 19 March 2019).
[1] Schmedtmann, J. and M. Campagnolo (2015), “Reliable Crop Identification with Satellite Imagery in the Context of Common Agriculture Policy Subsidy Control”, Remote Sensing, Vol. 7/7, pp. 9325-9346, http://dx.doi.org/10.3390/rs70709325.
Notes
← 1. In particular, the CAP aims to:
“support farmers and improve agricultural productivity, so that consumers have a stable supply of affordable food
ensure that European Union (EU) farmers can make a reasonable living
help tackling climate change and the sustainable management of natural resources
maintain rural areas and landscapes across the European Union
keep the rural economy alive promoting jobs in farming, agri-foods industries and associated sectors” (European Commission, n.d.[6]).
← 2. And abstracting away from questions of additionality (i.e. whether farmers prefer to act in a way that is consistent with the policy even in the absence of the policy).
← 3. Here, we do not consider the overall cost-effectiveness of the CAP policy as a whole: such a consideration would require consideration of all relevant costs and benefits, not simply the administrative costs of the monitoring and control system and the benefits of maintaining effective standards of compliance.
← 4. For example, the checks by monitoring approach developed by the Commission as an alternative to traditional checks on-the-spot, the Horizon 2020 SEN4CAP project. See Box 6.2.
← 5. Draxis Environmental S.A. (Leader) (GRC), Instituto Navarro de Technologias e Infraestructuras Agroalimentarias SA (ESP), Payment and Control Agency for Guidance and Guarantee Community Aid, National Paying Agency of Lithuania (LTU), Viesoji Istaiga Lietuvos Zemes Ukio Konsultavimo Tarnyba (LTU), Strutt & Parker LLP (GBR), Inosens Doo Novi Sad (SRB), University of Reading (GBR), National Observatory of Athens (GRC), Iniciativas Innovadoras Sal (ESP), ETAM S.A. (GRC) and CREVIS SPRL (BEL).
← 6. This objective is related with the Software Development Kit (SDK). The SDK provides RECAP platform users (in particular agricultural consultants responsible for farmers registered in the platform) with tools that help them build new added-value services upon the RECAP platform. This functionality enables the use and reuse of open data. For example, agricultural consultants can retrieve data of the parcels declared, along with results derived by the RS component, for use in other applications.
← 7. The platform uses an open licence (GNU General Public License version 3; info available at: https://opensource.org/licenses/GPL-3.0). It is not intended to entail a cost for farmers. Certain costs relating to customisation or adaptations may be incurred by the paying agencies and other interested authorities. Operational costs are to be covered by the paying agencies.
The platform source code is available at: https://zenodo.org/record/1451796#.XOuNXYj7RPY
← 8. These requirements are: Greening 1—Crop Diversification; Greening 2—Permanent Grassland; GAEC1—Buffer Strips; GAEC 4—Minimum Soil Cover; GAEC 5—Minimising Soil Erosion; SMR 1—Reducing water pollution in Nitrate Vulnerable Zones (VNZs) and GAEC 6—Maintaining the level of organic matter in soil (sources: https://ec.europa.eu/agriculture/direct-support/cross-compliance_en and https://ec.europa.eu/agriculture/direct-support/greening_en, accessed August 2018).
← 9. The first iteration used data from June 2018, right after the completion of farmers’ applications; and refers to the classification performed using satellite imagery until late June 2018. The second iteration was in late August 2018; and refers to the classification performed incorporating additional imagery (new Sentinel-2 acquisitions) that was acquired throughout the summer.
← 10. For further details about the accuracy assessment, contact the case study participants.
← 11. “Green light” signifies an almost completely trustworthy decision, yellow a less reliable but still usable decision, and red and unreliable being decisions of low confidence (these should be used with caution).
← 12. The classification is performed for crop types (i.e. soft wheat, barley, oats), crop clusters/families (i.e. cereals, legumes, maize, etc.) and crop season (i.e. summer, winter, permanent). All three levels of crop classification are provided to the PAs. According to the Greek Paying Agency, most of cross compliance rules are decided based on the crop cluster (family), with the exception of Greening requirements that require the lowest level of crop type differentiation.
← 13. Case study participants reported that initially, RECAP was aiming to develop a platform with a common interface and features for the delivery of public services. However, based on the results derived from the users’ needs analysis and co-production of services in all pilot countries, the technical team designed and developed five different interfaces/workflows customised to the specific needs of each of the five countries’ users. (Source: Personal Communication. Case study participant, Dimitrios Petalios (CREVIS), June 2018)
← 14. See https://ec.europa.eu/agriculture/glossary/acquis-communautaire_en_en, accessed August 2018.
← 15. See https://www.recap-h2020.eu/inspire_2018_conference/, accessed August 2018.
← 16. In theory, geo-tagged photos could also be used by farmers in support of an appeal. However, the use of geo-tagged photos to make an appeal may require changes to the existing EU CAP administrative and legislative frameworks. Consideration of such changes are beyond the scope of this case study.
← 17. Note that the RECAP platform does not ensure that administrations will not ask farmers to provide information already obtained, but the principle is that information provided into the RECAP platform will be available; i.e. the platform allows for, but does not intrinsically ensure, that the “once only” principle is implemented.