D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems

Towards a sustainable Open Data ECOsystem D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems This project has received funding from the European Unionʼs Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 955569. The opinions expressed in this document reflect only the authorʼs view and in no way reflect the European Commissionʼs opinions. The European Commission is not responsible for any use that may be made of the information it contains. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems Project Acronym ODECO Project Title Towards a sustainable Open Data ECOsystem Grant Agreement No. 955569 Start date of Project 01-10-2021 Duration of the Project 48 months Deliverable Number D5.1 Deliverable Title Models of allocating roles, tasks and responsibilities in open data ecosystems Dissemination Level Public Deliverable Leader Centre National de la Recherche Scientifique (CNRS) Submission Date 28-11-2024 Authors Ramya Chandrasekhar (CNRS), Davide Di Staso (TU Delft|), Melanie Dulong de Rosnay (CNRS) Co-author(s) María Elena López Reyes (AAU) Giorgos Papageorgiou (FAROSNET S.A.) Alejandra Celis Vargas (AAU) Liubov Pilshchikova (TU Delft) Caterina Santoro (KUL) Héctor Ochoa Ortiz (UNICAM) Ashraf Shaharudin (TU Delft) Umair Ahmed (UNICAM) Abdul Aziz (UNIZAR) Dagoberto Herrera (UNIZAR) Mohsan Ali (UAEGEAN) Maria Ioanna Maratsi (UAEGEAN) Document history Version # Date Description (Section, page number) Author & Organisation 20-06-2024 Kick-off meeting Melanie Dulong de Rosnay (CNRS) 26-08-2024 First brainstorming meeting for workshop to be conducted at TW5 Davide Di Staso (TU Delft), Ramya Chandrasekhar (CNRS) 30-08-2024 Workshop at TW5 Facilitators: Davide Di Staso (TU Delft), Ramya Chandrasekhar (CNRS) Participants: María Elena López Reyes (AAU) Giorgos Papageorgiou (FAROSNET S.A.) Alejandra Celis Vargas (AAU) Liubov Pilshchikova (TU Delft) Caterina Santoro (KUL) Héctor Ochoa Ortiz (UNICAM) Ashraf Shaharudin (TU Delft) Umair Ahmed (UNICAM) Abdul Aziz (UNIZAR) Dagoberto Herrera (UNIZAR) Mohsan Ali (UAEGEAN) Maria Ioanna Maratsi (UAEGEAN) D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems Version # Date Description (Section, page number) Author & Organisation 14-10-2024 Feedback from ESRs on case-studies María Elena López Reyes (AAU) Giorgos Papageorgiou (FAROSNET S.A.) Alejandra Celis Vargas (AAU) Liubov Pilshchikova (TU Delft) Caterina Santoro (KUL) Héctor Ochoa Ortiz (UNICAM) Ashraf Shaharudin (TU Delft) Umair Ahmed (UNICAM) Abdul Aziz (UNIZAR) Dagoberto Herrera (UNIZAR) Mohsan Ali (UAEGEAN) Maria Ioanna Maratsi (UAEGEAN) V0.1 16-10-2024 First draft, submitted for internal peer review Ramya Chandrasekhar (CNRS), Davide Di Staso (TU Delft), Melanie Dulong de Rosnay (CNRS) V0.1a 20-10-2024 First Peer review Joep Crompvoets (KU Leuven) V0.2 08-11-2024 Second draft. Processed feedback obtained from the peer review process Ramya Chandrasekhar (CNRS), Davide Di Staso (TU Delft), Melanie Dulong de Rosnay (CNRS) V0.2a 10-11-2024 Second peer review Joep Crompvoets (KU Leuven) V0.2b 08-11-2024 Feedback from WP Leader Charalampos Alexopoulos, UAEGEAN V0.2c 13-11-2024 Peer review Bastiaan van Loenen (TU Delft) V0.3 15-11-2024 Third draft. Processed feedback obtained from peer reviewers and work package leader Ramya Chandrasekhar (CNRS), Melanie Dulong de Rosnay (CNRS) V0.4 22-11-2024 Final review by scientific project coordinator Bastiaan van Loenen (TU Delft) V0.5 27-11-2024 Final report Ramya Chandrasekhar (CNRS), Davide Di Staso (TU Delft), Melanie Dulong de Rosnay (CNRS) V0.6 27-11-2024 Approval Bastiaan van Loenen, TUD V1.0 28-11-2024 Final editing Danitsja van Heusden, TUD D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems Table of Contents Abbreviations ..................................................................................................................................................................... 6 1 Executive summary ................................................................................................................................................. 7 2 Introduction ............................................................................................................................................................... 8 3 Methodology ..........................................................................................................................................................10 3.1 Literature review .......................................................................................................................................10 3.2 Workshop ....................................................................................................................................................10 3.3 Post-workshop data collection ...........................................................................................................13 3.4 Limitations of the methodology .........................................................................................................13 4 The Institutional Analysis and Design Framework ...................................................................................14 4.1 What is the Institutional Analysis and Design Framework ......................................................14 4.2 Benefits of the IAD Framework ...........................................................................................................15 4.3 How to apply the IAD Framework .....................................................................................................15 5 Applying the IAD Framework to an open data ecosystem – Contextual information ...............17 5.1 Background environment .....................................................................................................................17 5.2 Resource (socio-technical) attributes ...............................................................................................18 5.3 Actors ............................................................................................................................................................18 5.4 Rules-in-use and general governance structure ..........................................................................19 5.4.1 Rules-in-use with regard to openness of the resource and of the community .........19 5.4.2 General governance structure(s) ..................................................................................................20 5.4.3 Specific rules and norms ..................................................................................................................20 5.5 Goals and objectives ...............................................................................................................................20 6 Applying the IAD Framework to an open data ecosystem - Action arena .....................................23 6.1 Case-study #1: Breakthrough Open Data Project, Netherlands ............................................24 6.2 Case-study #2: Open Access Tax Benefit Micro Simulation Model (beamm.brussels), Belgium 26 6.3 Case-study #3: Gieß den Kiez project, CityLAB Berlin, Germany ...........................................28 6.4 Case-study #4: Collaborative investigation on femicides in Europe, European Union 30 6.5 Summary of all case-studies ................................................................................................................32 7 Applying the IAD Framework to an open data ecosystem – Evaluating patterns of interaction 34 7.1 Findings from case-studies ..................................................................................................................34 7.1.1 Summary of findings .........................................................................................................................35 7.1.2 Specific findings on roles and resources for non-government data holders .............36 7.1.3 Observations on goals and objectives of open data initiatives .......................................36 7.1.4 Observations on benefits and challenges of collaboration ...............................................37 7.2 Findings on institutional factors for sustainable collaborations in the open data ecosystem .....................................................................................................................................................................38 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 8 Conclusions .............................................................................................................................................................40 8.1 Objective of Task 5.1 ...............................................................................................................................40 8.2 Detailed conclusions ...............................................................................................................................40 8.3 Future work .................................................................................................................................................42 9 References ................................................................................................................................................................44 9.1 Literature......................................................................................................................................................44 9.2 Previous ODECO reports .......................................................................................................................46 9.3 Legal instruments of the European Union .....................................................................................46 Annex 1: Copies of Pitch Sheets and RACI Tables prepared during the workshop conducted at TW 5, Samos .............................................................................................................................................................................47 List of figures Figure 1: Collage of pictures from workshop conducted at ODECO TW 5, Samos ..............................11 Figure 2: Screenshot of presentation at the workshop conducted at ODECO TW 5, Samos, showing the variables that participants had to work with ...............................................................................................11 Figure 3: RACI sheets produced by participants during the workshop at ODECO TW 5, Samos ...13 Figure 4: IAD Framework. Source: Frischmann, Strandburg and Madison, 2014, pp 26. Green boxes are authorsʼ own additions. ........................................................................................................................................14 Figure 5: Questionnaire for using the IAD Framework ....................................................................................16 Figure 6: IAD Framework, with emphasis on Resource Characteristic, Attributes of the Community and Rules-in-Use. ...........................................................................................................................................................17 Figure 7: Authorsʼ visualisation of roles discharged by open data user groups, based on previous ODECO reports ................................................................................................................................................................19 Figure 8: IAD Framework, with emphasis on Action Arena. ..........................................................................23 Figure 9: IAD Framework, with emphasis on Patterns of Interactions. ......................................................34 List of tables Table 1: Pitches produced by participants at the workshop (see also Annex 1) ...................................12 Table 2: Synthesis of open data user needs, from previous ODECO reports ..........................................21 Table 3: Case-studies of collaborations in the open data ecosystem ........................................................24 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 6 Abbreviations D Deliverable DoA ODECO Description of Actions ESR Early Stage Researcher ESRI Environmental Systems Research Institute, Inc IAD Institutional Analysis and Design M Milestone NGO Non-governmental organisations ODECO Open Data ECOsystem ODECO TW 5 ODECO Training Week 5 RACI Responsible, Accountable, Consulted, Informed – Hierarchy of tasks T Task WP Work Package Nr Partner Partner short name Country Beneficiary 1 Technische Universiteit Delft TU Delft Netherlands 2 Katholieke Universiteit Leuven KUL Belgium 3 Centre National de la Recherche Scientifique CNRS France 4 Universidad de Zaragoza UNIZAR Spain 5 Panepistimio Aigaiou UAEGEAN Greece 6 Aalborg Universitet AAU Denmark 7 Università degli Studi di Camerino UNICAM Italy 8 Farosnet S.A. FAROSNET S.A. Greece Partner organisations 1 7eData 7EDATA Spain 2 Digitaal Vlaanderen DV Belgium 3 City of Copenhagen COP Denmark 4 City of Rotterdam RDAM Netherlands 5 CoC Playful Minds CoC Denmark 6 Derilinx DERI Ireland 7 ESRI ESRI Netherlands 8 Maggioli S.p.A MAG Italy 9 National Centre of Geographic Information CNIG Spain 10 Open Knowledge Belgium OKB Belgium 11 SWECO SWECO Netherlands 12 The government lab GLAB United States of America 13 Agency for Data Supply and Infrastructure ADSI Denmark 14 GFOSS Open Technologies Alliance GFOSS Greece 15 Inno3 Consulting IC France 16 Regione Marche RM Italy 17 Open Data Institute ODI United Kingdom 18 Swedish National Archives SwNA Sweden D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 7 1 Executive summary This Task 5.1 is focussed on the potential impact of different forms of collaborations (including public-private partnerships) to sustainably arrange the distribution of roles, tasks and resources in open data ecosystems. Building on previous ODECO research on governance of open data ecosystems and drawing from a participative workshop conducted during ODECO Training Week 5 held at Samos, Greece, this report summarises the distribution of roles, tasks and resources in four open data case-studies – the Breakthrough Open Data Project, beamm.brussels, CityLAB Berlin trees project, and Femicides in Europe. Each case-study serve as an illustrative model of collaboration between a public administration and one or more open data actor group. By applying the Institutional Analysis and Design Framework (a preferred tool for policy analysis) to each case-study, this report summarises a set of institutional factors that could be relevant for sustaining and replicating models of collaboration in other open data ecosystems. These institutional factors are: • There should be continued focus on publication of open government datasets by public administration. • Coordinating functions should be discharged by public administrations to bring together open data providers and open data users to openly discuss their needs. In some cases, such coordinating functions can be discharged by other non-governmental stakeholders as well. • Partnerships should be formed between public administrations / international organisations and other civic open data user groups to co-develop open data initiatives, which can prompt contributions from other actors. • Shared open infrastructures for publishing and use of data and information are important, including for non-governmental data holders. • A strong interorganisational culture oriented towards effecting social and economic impact of open data in a participative manner is also useful. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 8 2 Introduction This Task 5.1 is focussed on the potential impact of different forms of collaborations (including public-private partnerships) to sustainably arrange the distribution of roles, tasks and resources in open data ecosystems. According to the DoA, this task “will design and review different models of allocating roles, tasks and resources in open data ecosystems. ESR1 will identify different roles for actors to participate in the open data ecosystem and explore practices of participatory design to allow actors to actively participate in and contribute to the open data ecosystem. The design will build on the perspective of the eight actor groups: non-specialist data users (ESR1), local government (ESR6), journalists (ESR9), students (ESR10), NGOs (ESR11), central/regional government (ESR12), companies (ESR13), and data intermediaries (ESR15) as presented in the results of T2.1, T2.3, T3.3 and T4.3. CNRS (ESR4: forms of public-private partnerships), AAU (ESR6: local government perspective), KULEUVEN (ESR12: central government perspective) and UNICAM (ESR13: companiesʼ perspective) will explore the potential and impact of different forms of public-private partnerships as a means to sustainably arrange for the distribution of roles and tasks in the ecosystem." This report builds on previous ODECO research relating to open data governance – specifically reports prepared for Task 2.1 (Open Data User Needs: Seven Flavours), Task 2.3 (User needs from a governance perspective), Task 3.3 (Closing the cycle: Promoting open data usersʼ contribution from a governance perspective) and Task 4.3 (An approach to steer the behaviour of nongovernment data holders towards open data from a governance perspective). Further, this report is focussed on eight actor groups: non-specialist data users, local government, journalists, students, NGOs, central/regional government, local government, and open data intermediaries. In order to identify the different roles for actors to participate, we undertook a review of these previous ODECO outputs on open data governance that identified some roles, tasks and resources in open data ecosystems. We synthesised findings on the roles discharged by the identified open data actor groups in an open data ecosystem, and the types of value contributions made by these actor groups. Then, using practices of participatory design, we designed and conducted a workshop at ODECO TW 5, in Samos (Greece). This workshop produced real-life examples of collaboration between open data actorsʼ groups to create value from open data that serve as case-studies for this report. For each case-study, workshop participants identified specific tasks performed by each actor group, ranked these tasks in accordance with the degree of responsibility borne by each actor group, and identified governance instruments that facilitated the distribution of these tasks. This allowed us to identify the connections between the roles, tasks and resources, to find out if the knowledge had to be updated since the previous reports, to identify the various ways in which the actors group collaborate, and in which the resources are obtained and managed by those groups. Finally, we used the Institutional Analysis and Design framework, a theoretical framework suitable to represent the relations between resources and communitiesʼ attributes, rules and patterns of interactions, in a version adapted to knowledge resources (Frischmann, Madison, & Strandburg, 2014). The choice of this analytical framework was discussed and validated during the kick-off meeting for this Task 5.1 (held on June 20, 2024), as allowing to both represent the dynamics, analyse and evaluate the case-studies contributed during the ODECO TW 5 at Samos as well as D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 9 identify the potential and impact of the different types of collaborations (including public-private partnerships). The Institutional Analysis and Design framework is suitable for policy analysis. It has been used for analysing and evaluating collaborations for sustainable resource management in the context of shared information resources. We present some suggestions on institutional factors relevant for collaborations in open data ecosystems, which can aid in sustainable distribution of roles and tasks, as well as enable replicability of existing models of collaboration. In particular, we suggest five institutional factors to sustain collaborations and successfully govern open data ecosystems: • There should be continued focus on publication of open government datasets by public administration. • Coordinating functions should be discharged by public administrations to bring together open data providers and open data users to openly discuss their needs. In some cases, such coordinating functions can be discharged by other non-governmental stakeholders as well. • Partnerships should be formed between public administrations / international organisations and other civic open data user groups to co-develop open data initiatives, which can prompt contributions from other actors. • Shared open infrastructures for publishing and use of data and information are important, including for non-governmental data holders. • A strong interorganisational culture oriented towards effecting social and economic impact of open data in a participative manner is also useful. This report is part of Work Package 5 (Towards a Sustainable Open Data Ecosystem) that seeks to integrate all ODECO research until date towards design, policy and governance recommendations for sustainable open data ecosystems. This report for Task 5.1 outlines actor roles, tasks and responsibilities in open data ecosystems, and proposes a set of institutional factors that could aid in sustainable arrangement of these roles, tasks and responsibilities. These findings are grounded in the previous reports which had identified tasks, actors, resources and roles allocation, and have been updated with the contribution of the ESRs. This report, therefore, constitutes a gateway and an input designed to further feed into Task 5.2 on balancing and redistributing value in a sustainable open data ecosystem, and into Task 5.3 on the design of an open data ecosystem. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 16 Figure 5: Questionnaire for using the IAD Framework In this report, information relating to the background context, resource (technical) attributes, actors/community members, rules-in-use / governance practices, and goals and objectives are synthesised from a review of previous ODECO deliverables – specifically T2.1, T2.3, T3.3 and T4.3. Information relating to patterns of interaction were synthesised collaboratively by all ESRs, through a workshop conducted at TW5 and through post-workshop data collection. All this information is then cohesively analysed, to arrive at institutional recommendations for sustainable distribution of roles, tasks and resources within the selected open data ecosystem, to meet the objectives of T5.1. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 17 5 Applying the IAD Framework to an open data ecosystem – Contextual information In this section we describe certain features of an open data ecosystem (highlighted in blue in Figure 6 below), based on a literature review of previous ODECO deliverables. This serves to contextualise interactions within this ecosystem, and aids in evaluating whether collaborative interactions within this ecosystem can be replicated. Figure 6: IAD Framework, with emphasis on Resource Characteristic, Attributes of the Community and Rules-in-Use. 5.1 Background environment Frischmann, Strandburg and Madison (2014, pp 21) identify two types of cultural environments in the context of informational resource – the ‘naturalʼ cultural environment where informational resources are excluded from proprietary intellectual property regimes, and the ‘proprietaryʼ environment where informational resources (like code, datasets or databases) are subject to proprietary regimes like copyright that regulate and/or control access, sharing and use of these resources. For open data, the background cultural environment is the latter – the proprietary environment. In the European Union, both software code and databases are protected by copyright (EU Directive 2001/29/EC, Article 1). Further, databases are also protected under a sui generis framework, where substantial investment in “obtaining, verification or presentation of the contents to prevent extraction and/or re-utilization of the whole or of a substantial part of the contents of a database” (EU Directive 96/9/EC, Article 7). Open data become shared informational resources, i.e. resources without legal access or use restrictions, through the use of open data licenses where the resource owner/creator provides pre-facto authorisations or waives copyright/sui generis claims (Giannopoulou, 2018). In the context of open government data, legal regulations require that public sector bodies exercise their intellectual property rights over databases in a manner that enables re-use (see for e.g., EU Directive 2019/1024 Open Data Directive, Recital 54; EU Regulation 2022/868 Data Governance Act, Recitals 17 and 18). However, as De Filippi and Maurel (2015) note, there are underlying conflicts between legal regulations that encourage re-use of public sector information and the legal system of intellectual property that allow public sector bodies to assert a de-facto exclusive rights on public sector information. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 18 5.2 Resource (socio-technical) attributes Open data is digital data that is capable of free and unrestricted usage, sharing, and access to data in any available format (ODECO 2023b; Open Definition 2.1). From a technical perspective, open data comprises of multiple components such as individual data elements, datasets that constitute aggregated data elements, databases which contain multiple datasets, platforms that house and/or provide access to databases, metadata, and open data standards. From a functional perspective, open data is not merely data that is ‘obtainableʼ by anyone. Rather, it is data that is: (i) available as a whole in a convenient and modifiable form, (ii) provided under terms that permit its reuse and redistribution, including ‘remixingʼ with other data, and (iii) subject to universal participation in its use, reuse and redistribution (Dove 2015, pp 158). Further, unlike natural resources, open data is not naturally occurring but is brought into existence. As a result, the attributed of open data also include the human labour and choices involved in its production and maintenance as well as the material artefacts necessary for its production and use such as sensors, data centres, cables, computers, etc. And finally, as scholars of critical data studies and STS note, data is not only an object/economic resource but also a process of power. Sociopolitical decisions about what data to generate (and what not to) as well as what to do with data are inherently linked to the technical attributes of open data (Gurstein, 2011; Kitchin and Lauriault, 2014; Fisher and Streinz, 2022). Beyond open data, ODECO (2024b) identifies certain other value contributions made to an open data ecosystem. These include outputs generated from open data (such as data analysis, visualisations, data stories), insights generated from open data (such as local knowledge and social issues of concern) and technological innovations to enable the use of open data (such as websites, platforms, applications, tools). 5.3 Actors Pursuant to the ODECO DoA, this report incorporates the perspectives of 8 open data actor groups: • Non-specialist data users • Local government • Journalists • Students • NGOs • Central/regional government • Companies • Open data intermediaries From existing ODECO reports (2023a, 2024b, 2024c), it is also possible to deduce the different roles that these actors perform in an open data ecosystem: D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 19 Figure 7: Authorsʼ visualisation of roles discharged by open data user groups, based on previous ODECO reports 5.4 Rules-in-use and general governance structure Frischmann, Strandburg and Madison (2014, pp 28) recommend an analysis of rules-in-use at three levels – the degree of openness of the resource and of the community, the general governance structure, and the specific rules and norms that apply to specific action arenas. 5.4.1 Rules-in-use with regard to openness of the resource and of the community ODECO (2023a; 2024a) identified certain rules-in-use that serve as barriers to use open data. Definitionally, open data is meant to available for unrestricted re-use. In practice, there are different legal modalities by which data is made available as open data – through the voluntary use of open data licenses, or by virtue of legal mandates such as the EU Open Data Directive to publish public sector information in machine-readable formats (ODECO 2024a, pp 34). There are also sectoral legal regulations for release of certain datasets to the public, such as Article 11(1) of the EU INSPIRE Directive which requires public bodies to make spatial datasets available to the public, and the USʼs Equitable Data Collection and Disclosure on COVID-19 Act which required the US federal government to collect and publicly release racial and other demographic data on COVID-19 (Id.) To obtain data from the private sector, public administrations sometimes do rely on data-for-data agreements. For example, one of the respondents to a questionnaire circulated for Task 2.3 referred to data-for-data arrangements between the Netherlands Transport Department and Vodafone, to obtain data on traffic intensity from Vodafone, combined with other data held by the Transport Department, and to release all such data as open data by the Transport Department. Even within licenses, there is a lack of standardization (ODECO 2023a, 2024a). Further, there are technical barriers to the use of open data, owing to challenges in translating the FAIR (Findable, Accessible, Interoperable, Reusable) principles into practice (ODECO 2023b). For D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 20 instance, open data portals often have only one input method for search and should include alternative inputs methods to search the repository. Open data portals are also often not optimised for use via smartphones. ODECO (2024a) also identified certain rules-in-use with regard to open data actors. For instance, in some cases, certain open data actors come together to form either as communities of purpose (as in the case of OSM, where humanitarian actors come together with open source coders and mappers, to map conflict/disaster affected areas) or as communities of practice (as in the case of certain scientific data) (ODECO 2024a, pp 30-31). In each of these communities, membership rules are differently defined – through informal rules as in the case of communities of purpose as opposed to more formal membership criteria as in the case of communities of practice. 5.4.2 General governance structure(s) An ecosystemic perspective to open data moves beyond a linear understanding of open data as mere data disclosures by public bodies, and instead probes the different forces of competition and collaboration that sustain the generation and re-use of open data. From this perspective, ODECO has made significant contributions in three aspects: • ODECO (2024b; 2024c) has investigated what actors contribute as value. This includes contribution of open data by both government and non-government actors, but also other value contributions in form of knowledge outputs, technological innovations, advocacy, etc. • ODECO (2024a; 2024b; 2024c) has also investigated why certain actors contribute such value, by investigating their intrinsic and extrinsic motivations as well as existing governance instruments. • ODECO (2024c) has investigated how certain actors can contribute (more) open data, by investigating governance instruments necessary for such open data contributions. Drawing from Crompvoets et al (2019), ODECO (2024c) identified three broad mechanisms underpinning governance in an open data ecosystem – hierarchies, markets, networks. From this broad categorization, it is possible to identify the processes that underpin these three governance mechanisms, such as authority, price and competition or trust and solidarity. These processes, in turn, rely on specific governance instruments. ODECO (2024c) classifies these governance instruments into two types – structural and managerial. The analysis undertaken in ODECO (2024c) revealed that a mix of hierarchical measures, such as legal frameworks as well as grassroots network initiatives in the form of collaborative partnerships, can effectively stimulate open data sharing. However, market-oriented instruments, such as financial incentives, are currently underutilized, likely due to the non-commercial nature of many non-governmental actors we analysed in this study. These findings can be applied not only to the sharing of more open data, but also to the facilitation of other forms of participation from non-governmental actors. 5.4.3 Specific rules and norms Through the workshop conducted at TW5, ESRs identified certain case-studies of collaboration between open data actors. Upon further iteration, ESRs identified governance instruments that facilitate the distribution of tasks among the actors in each case-study. These governance instruments are examples of specific rules and norms and are described in more detail in Section 6 below. 5.5 Goals and objectives The benefits of open data include enhancing the effectiveness and efficiency of public services, increasing institutional accountability, boosting citizen participation, accelerating scientific progress, and fostering the creation of other economic and social values (ODECO 2023a, Hossain et al., 2016; Janssen et al., 2012; Zhu et al., 2019). D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 21 A review of existing ODECO reports on governance using interpretive analysis also reveals what kinds of value realised are by and/or contributed by the open data user groups through participation in an open data ecosystem. We refer to these collectively as ‘value typesʼ, and they comprise of the following: • The generation of more open data • Improving the quality of existing open datasets • Create technological innovations (platforms/apps) for the use of open data • Create businesses through open data • Create visualisations and knowledge from open data • Identify social issues of concern • Hold governments accountable • Create products from open data • Become informed citizens • Ensure transparency of public administrations • Provide funding for open data initiatives Some challenges to the realisation of these value types have also been identified in previous ODECO reports. One set of challenges relate to the incorporation of user needs into policy design for open data ecosystems. ODECO (2023a) identified some commons user needs across 9 user group types – local government, regional/central government, companies, open data intermediaries, artificial users, non-specialist data users, journalists, students, and nongovernmental organisations. The identified user needs can be grouped in three levels – user needs relating to resource attributes, user needs relating to actorsʼ competencies, and user needs relating to the broader open data ecosystem. Table 2: Synthesis of open data user needs, from previous ODECO reports User needs relating to resource attributes User needs relating to actorsʼ competencies User needs relating to the broader open data ecosystem Access to data, availability and findability Findability and metadata extraction of open data is a challenge, including adherence to the 5-Star Principles for Linked Open Data. Data availability is also not uniform across territories, nor equally representative of the concerns of different citizen groups. Literacy All 9 user groups require a broad range of skills to interact and participate in an open data ecosystem. These include data and digital literacy, critical and scientific thinking, and ethics. Funding The participation of users in an open data ecosystem depends on the amount of funding/economic capital made available for such participation. Data Quality This includes both technical quality of open datasets (i.e. their completeness, accuracy, findability and metadata quality) as well as the social quality (i.e. the manner in which datasets accurately reflect social realities) Data ethics Users must be aware of data ethics and privacy issues when making use of open data Regulation There is a need for greater regulatory clarity in terms of standard licensing for open data, enabling easy re-use. Further, regulation is also required to ensure equitable data access to facilitate the participation of diverse D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 22 User needs relating to resource attributes User needs relating to actorsʼ competencies User needs relating to the broader open data ecosystem groups and satisfy societal needs. Data infrastructure Digital open data require well-designed and wellmaintained IT and technical infrastructure. Communication Different usergroups have different communication needs. For example, local governments need to communicate with public bodies and private companies for the generation of open data, while NGOs need to communication with a range of partners to find open data. Governance and coordination There is a need for both organisational governance as well as data governance, to ensure inclusive data use practices. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 23 6 Applying the IAD Framework to an open data ecosystem - Action arena As mentioned in Chapter 3.2 above, we conducted a workshop at TW 5, Samos, where we used participative design principles to identify real-life examples of collaboration between actors in an open data ecosystem to realise different types of value. This section summarises each real-life example of collaboration as a ‘case-studyʼ (see Table 3). In terms of the IAD Framework, these case-studies represent the action situations within the open data ecosystem (highlighted in blue in Figure 8 below): Figure 8: IAD Framework, with emphasis on Action Arena. These case-studies illustrate: (i) different types of collaborations between actors in an open data ecosystem, and (ii) different types of value realised from such collaborations - either in the form of generation of more or new open datasets, or in the form of other social and/or economic benefits. Three specific insights emerge from these case-studies, which serve as the basis for evaluating whether they present sustainable forms of collaboration in the open data ecosystem: (i) how tasks are divided among actors involved in the ecosystem, (ii) what governance instruments facilitate the distribution of these tasks, and (iii) whether any institutional factors are relevant for the distribution of these tasks, replicability of the collaborations, and sustainability of the collaborations. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 24 Table 3: Case-studies of collaborations in the open data ecosystem # Title 1 Breakthrough Open Data Project, Netherlands 2 Open Access Tax Benefit Micro Simulation Model (beamm.brussels), Belgium 3 Gieß den Kiez project, CityLAB Berlin, Germany 4 Collaborative investigation on femicides in Europe, European Union Each case-study also contains the RACI Table prepared by ESRs during ODECO TW 5 at Samos and subsequently iterated by the ESRs. Each RACI Table identifies the actors involved in a case-study, the tasks performed by each actor, and their level of involvement each task. This was done by identifying four degrees of involvement: • R as in ‘responsibleʼ - the actor who actually executes the task, • A as in ‘accountableʼ - the actor reviews/supervises the task, • C as in ‘consultedʼ - the actor provides input on the task, • I as in ‘informedʼ - the actor who is kept in loop about the progress of the task. 6.1 Case-study #1: Breakthrough Open Data Project, Netherlands Description of case-study This case study is about the geospatial data sector in the Netherlands. The Dutch government initiated a project called ‘Breakthrough Open Geodata Projectʼ that ran from 2013 to 2017. The initiative was led by the CEO of Esri. The project comprised of representatives from businesses, public sector, and academia: GeoBusiness Nederland (trade association for companies that work with geoinformation), Esri Nederland, Ministry of Economic Affairs, Ministry of Infrastructure and the E nvironment, Ministry of Interior and Kingdom Relations, and Delft University of Technology. The project aimed to bring together societal demand, entrepreneurs who develop applications, and holders of open data through two pillars. The first pillar is to shift the supply-oriented approach of open data provision towards demand-driven approach. The second pillar is to ensure continuous provision of high-quality open data in order to provide certainty to market actors and facilitate solid business models based on open data. URL • https://adoc.pub/doorbraakproject-open-geodata-als-grondstof-voorgroei-en-inc207c91981124eda8cc49d9e21c39fbc80957.html • https://ibestuur.nl/artikel/doorbraakproject-open-geodata-kan-zonderdoorbraak/ Actors involved • GeoBusiness Nederland (representing interests of private sector users of open geodata) • Esri as an open data intermediary (that is also part of GeoBusiness Nederland) • Dutch national government • Delft University of Technology (TU Delft) (representing the academic community) Types of value contributions Tangible outcomes from the project include the release of actual elevation data (AHN) and data from the Dutch space office (NSO) as open data. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 25 Distribution of tasks among actors Actor: Dutch government Actor: GeoBusiness Nederland Actor: Esri Actor: TU Delft Actor: Other users of open geodata Task : Communicate interests I R R R C Task: Provide recommendations I R R R C Task: Publish open data R C C C I Task: Use open data I R R R R • GeoBusiness Nederland and Esri communicate the interests of private sector open geodata users to the government • TU Delft provides recommendations for open geodata strategies • Government agencies release actual elevation map (AHN) and satellite data from the Dutch Space Office as open data • Other open geodata users benefit from the initiative: e.g., researchers from Wageningen University reported that solar panels can be installed more efficiently by using AHN in combination the Basic Registration of Addresses and Buildings (BAG) (that is also open data ), and archaeologists use AHN data to locate ancient settlements that are not noticeable to the naked eye Governance instruments facilitating distribution of tasks Structural instrument: • Establishment of coordinating functions or entities Observations on institutional factors enabling collaboration • The collaboration was initiated by the government but was led by a major player in the geoinformation industry, namely Esri. • While this project exemplifies the potential form of coordination that could take place to improve the provision of open government data (i.e., user-driven), it had limited impacts in making the non-public sector share open data (i.e., inclusivity) and in improving the distribution of costs and benefits within the open data ecosystem (i.e., circularity). D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 32 6.5 Summary of all case-studies In Table 3 below, we summarise the distribution of tasks from the 4 case-studies. Specifically, we highlight the division of tasks between governmental and non-governmental actors. Table 3: Summary of division of tasks between governmental and non-governmental actors, derived from 4 case-studies Name of casestudy Tasks performed by governmental actors Tasks performed by non-governmental actors Regional and national government Local government Academic institution Industry Journalists NGO Citizens as Nonspecialised users Breakthrough Open Data Project • Responsible for publishing open geospatial data • Informed about and acted on needs of open data users, and informed about use of open data N/A • Advised the regional governme nt on open data strategies • Communic ated needs and interests of private sector actors regarding availability and use of open data N/A N/A N/A beamm.brussel s • Both government s are responsible for publishing statistical data • Regional government is responsible for providing funding for the initiative • Both government s are informed of data analysis and the simulation N/A • Responsibl e for creation of synthetic datasets from existing data (including statistical data), developin g the tax similar model, and verifying the model • Academic institution must remain accountabl N/A N/A N/A • Informed of the tax simulation tool as potential users of the tool. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 33 and verification of the tax simulation model e for funding received from regional governme nt CityLAB Berlin • Responsible for publishing open datasets on tree registry in Berlin • Responsible for publishing open datasets on tree registry in Berlin N/A N/A N/A • Responsibl e for collecting data from citizens and combining with open data • Also responsibl e for creating and maintain the Gieß den Kiez website • Responsibl e for providing crowsourced data on tree health • Consulted on creation of the Gieß den Kiez website Femicides in Europe • Regional/nat ional government s (together with the European Union) were accountable for creating statistical data on violence against women N/A N/A N/A • Responsibl e for collecting and combining data on violence against women • Responsibl e for creating visualisatio ns • Responsibl e for collecting additional data, maintainin g the datasets, and verifying obtained data • Consulted by journalists when collecting and preparing datasets on violence against women • Responsibl e for creating visualisatio ns • Responsibl e for obtaining additional data and verifying obtained data • Informed of the database and visualisatio ns D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 34 7 Applying the IAD Framework to an open data ecosystem – Evaluating patterns of interaction In this section, we combine the contextual information on open data ecosystems (derived through literature review of existing ODECO reports) with the real-life examples of collaborations in an open data ecosystem (derived through a participative design-based workshop). From the perspective of the IAD Framework, we combine these aspects to unearth insights on patterns of interactions (highlighted in Figure 9 below), to analyse whether the collaborations we studied can be replicated. We also extract and synthesise some observations on institutional factors that influence the distribution of roles, tasks and resources in open data ecosystems in a sustainable manner. Figure 9: IAD Framework, with emphasis on Patterns of Interactions. 7.1 Findings from case-studies Per the IAD framework as applied to shared informational resources, the purpose of evaluating different action situations in accordance with certain pre-defined metrics is to “extract generalizable knowledge that will be useful for policy makers” (Frischmann, Madison and Strandburg, 2014, pp 36). In this report, we analysed the four case-studies described in Section 6 above as the action situations within an open data ecosystem. While these case-studies are not representative of all types of collaborative interactions between open data actor groups to realise value from open data nor are they representative of all types of open data ecosystems, they nonetheless constitute empirical data from which insights on the success and replicability of these interactions can be gleaned. The IAD Framework requires that patterns and outcomes stemming from the action arena be evaluated in terms of: (i) the solutions and benefits they offer for a particular collective action problem, and (ii) the associated costs and risks (Frischmann, Madison and Strandburg, 2014, pp 37). Accordingly, we evaluate the four case-studies summarised in Chapter 6 above to assess: (i) what goals and objectives they valorise, and (ii) what challenges and benefits they bring to the realisation of social and economic value from open data. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 35 7.1.1 Summary of findings All four case-studies are examples of collaborations between public administrations (either at local, regional or national level) and one or more open data user-groups. In one case-study (Breakthrough Open Data Project), a nongovernment collaborators comprised of industry and academia. In one case-study (beamm.brussels), the primary nongovernment collaborator was an academic institution. In one case-study (CityLAB Berlin), the nongovernment collaborators were an NGO and citizens. In another case-study (Femicides in Europe), the nongovernment collaborators were NGOs and journalists. Three of the case-studies (beamm.brussels, CityLAB Berlin and Femicides in Europe) involved citizens as non-specialised users. All 4 case-studies serve as examples of open data initiatives being jointly designed and implemented by public administrations with the primary collaborators. In this regard, these casestudies constitute examples of collaboration within an open data ecosystem – i.e. an ecosystem consisting of the 8 actor groups mentioned in the DoA, where these actors not only make use of open data for their own purposes, but also contribute value back in different ways. In terms of value-contributions, one of the case-studies (Breakthrough Open Data Project) resulted in the release of existing datasets as open data. One of the case-studies (CityLAB Berlin) resulted in the production of new open datasets. One case-study (Femicides in Europe) sought to improve the quality of existing open datasets. Two case-studies (beamm.brussels and CityLAB Berlin) contributed new visualisations, to create knowledge from open data. Three case-studies (beamm.brussels, CityLAB Berlin and Femicides in Europe) identified social issues of concern from the open data generated/analysed. And finally, two case-studies (CityLAB Berlin and Feminicides in Europe) sought to create a civic community around the social issue identified. In terms of distribution of tasks, two of the case-studies (Breakthrough Open Data Project and beamm.brussels) relied on public administration (either at regional or national levels) to release open datasets. In Breakthrough Open Data Project, the focus was two-fold: (i) to shift the supplyoriented approach of open data provision towards demand-driven approach, and (ii) to ensure continuous provision of high-quality open data in order to provide certainty to market actors and facilitate solid business models based on open data. In beamm.brussels, existing open datasets (such as census data) were relied on to create the synthetic dataset from which visualisations of fiscal policies were created. In one case-study (beamm.brussels), the academic collaborator was also identified as a responsible party to generate and release open datasets. In two case-studies ( CityLAB Berlin and Femicides in Europe), the collection and publication of open datasets was the responsibility of NGOs (in both cases) and journalists (in the case of Femicides in Europe). But in both these as well case-studies, the NGOs relied on and combined existing open datasets with other generated data. In three case-studies, certain actors were consulted in the generation and release of open datasets. In the Breakthrough Open Data Project, Esri and GeoBusiness Netherlands provided recommendations to the Dutch national government on what datasets should be provided as open data. In CityLAB Berlin, since the data on the watering of the trees was crowdsourced, the local Berlin government was consulted to update these data with available information on city watering schedule. In Femicides in Europe, other NGOs as well as the regional governments of the countries surveyed were consulted in the data generation, publication and visualisation processes. In terms of structural governance instruments, three case-studies (beamm.brussels, CityLAB Berlin and Femicides in Europe) identified the same structural instruments relevant for the distribution D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 36 of tasks – the creation of systems for sharing of information, and the existence of partnerships between different actors. Two case-studies (beamm.brussels and Breakthrough Open Data Project) flagged another strategic instrument – the establishment of coordinating functions or entities. In terms of managerial governance instruments, one case-study (beamm.brussels) identified strategic planning as a relevant instrument. In this case, the collaboration between CAPE and the Brussels regional government was under a framework which specified clear objectives and targets for the actors. Two case-studies (CityLAB Berlin and Femicides in Europe) flagged the importance of a vibrant interorganizational culture and knowledge management towards the creation of open data and knowledge and towards the creation of civic communities centred on social issues of concern. One case-study (CityLAB Berlin) flagged the importance of capacity building as a managerial governance instrument, to involve non-specialised users. Two case-studies (beamm.brussels and Femicides in Europe) flagged the importance of financial management for the sustained collaboration between a university and regional government (in the case of beamm.brussels) and between journalists, NGOs and regional governments (in case of Femicides in Europe). 7.1.2 Specific findings on roles and resources for non-government data holders All 4 case-studies involve participation of at least one non-governmental actor. In the Breakthrough Open Data Project, the non-government actors were GeoBusiness – an association of geo companies, Esri, and Delft University of Technology representing the academic community. These actors, together with the public sector, collectively advised on the need for more open geodatasets. In beamm.brussels, the non-governmental actor was CAPE – an academic research centre. This research centre created the synthetic datasets used to build the tax simulation model – by combining open datasets on statistics with other data on socio-economic indicators. CAPE shares these synthetic datasets upon request. Here, the non-governmental actor played the role of ‘open data userʼ as well as ‘data creatorʼ, albeit the datasets created by CAPE are not released as openly licensed datasets. In both CityLAB Berlin and Femicides in Europe, NGOs were responsible for data collection and for release of this data as open datasets. In Femicides in Europe, journalists contributed to this objective as well. As a result, these two case-studies reveal the roles discharged by NGOs and journalists as creators and holders of non-government open data. Three case-studies – beamm.brussels, CityLAB Berlin and Femicides in Europe – reveal the importance of shared systems for information exchange as a resource for these non-governmental actors to contribute open data as well as create value from open data. Two case-studies – beamm.brussels and Femicides in Europe – also reveal the importance of financial instruments in the form of funding, to enable the non-government actors to participate in these collaborations. 7.1.3 Observations on goals and objectives of open data initiatives All case-studies reveal that the creation of economic and social value can be achieved through collaborations between public administrations and more than one open data user group (such as academic institutions, journalists, open data intermediaries, NGOs and/or non-specialised users). One case-study (CityLAB Berlin) also reveal how environmental value can be realised from collaborations – in this case, the watering of trees in the city of Berlin. Two case-studies (Breakthrough Open Data Project and beamm.brussels) reveal the impact of the collaborations on enhancing effectiveness and efficacy of public services – in the former, enabling D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 37 the release of existing datasets as open data, and the in latter in improving the ability of policymakers to assess proposed fiscal policies. Three case-studies (beamm.brussels, CityLAB Berlin and Femicides in Europe) also reflect the ways in which collaboration with non-specialised users boosted citizen participation – by co-identifying social issues of concern as well as creating civic communities around these social issues. One of the case-studies (CityLAB Berlin) also serves as an example of open data collaboration enhancing scientific progress, by tracking tree watering in Berlin. 7.1.4 Observations on benefits and challenges of collaboration Frischmann, Madison and Strandburg (2014, pp 37-38) write that under the IAD framework, the costs and benefits of producing open data and other outputs/outcomes from open data based on the case-studies in the action arena should be compared against the production of similar resources/outputs/outcomes under “IP-based market transactions, government subsidy, or some alternative system.” In the Breakthrough Open Data Project, Esri and TU Delft provide recommendations on what types of geodata are necessary for users of such open data. These recommendations help improve the quality and quantity of open datasets published by the Dutch government. Legal frameworks like the Open Data Directive impose certain obligations on public administrations to release highquality geographical datasets. But in the absence of a channel of communication between the data provider (i.e. Dutch government) and data users (such as Esri, TU Delft and other geodata users), the legal framework requiring publication of open geodatasets by government bodies does not in-and-of-itself allow for a shift from a supplier-driven open data ecosystem to a user-driven open data ecosystem. On the other hand, in terms of costs, the Breakthrough Open Data Project was initiated by the Dutch government to improve provision of open government datasets, identifying and resolving barriers towards the use of open data, and recommending ways to improve the interaction between providers and users. It had limited impacts in making the nonpublic sector share open data (i.e., inclusivity) and in improving the distribution of costs and benefits within the open data ecosystem (i.e., circularity). In beamm.brussels, the collaboration between public administration and an academic institution to develop a simulation tool to assess fiscal policies illustrates how social equity can be achieved in policymaking. This enables policymakers to ‘seeʼ the potential impacts of fiscal policies before they are introduced, as opposed to after. Further, the existence of open datasets on tax and socioeconomic data about Brussels Region enabled the creation of this tool. As a result, it is important to ensure quality, accuracy and representativeness of these datasets. But the beamm.brussels project is not focused on this type of feedback loop between the data providers (i.e. Belgian regional governments) and the data user (i.e. UCLouvain as the creator of the simulation tool). In the CityLAB Berlin project, the collaboration between an NGO (CityLAB), citizens and local government enabled the creation of a civic community around the preservation of trees and green cover in the city of Berlin. This project integrates citizen-contributed data on tree watering with existing open datasets on tree cover, to improve the quality of existing open datasets. It also enables the creation of a civic culture around citizen contribution and use of open data. Further, the NGO plays an important role as a facilitator of citizen participation. Citizens could have created this project on their own, using existing open datasets, but this does not account for technical skill D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 38 gaps among citizens in locating open data and building technical tools like apps using open data. The NGO is an important actor in addressing these skill gaps. Finally, in the Femicides in Europe project, a multi-actor collaboration between journalists, NGOs, the European Institute for Gender Equality, regional governments, the European Union and citizens as non-specialised users, was crucial in drawing attention to the social issue of femicide in Europe. Funding provided by the European Union was crucial in sustaining this collaboration. Further, the presence of a shared interorganisational culture on the realisation of social value from open data also helped sustain the collaboration between diverse actors. In the absence of these factors, a collaboration of this scale could be possible in theory but would be difficult to execute in practice. 7.2 Findings on institutional factors for sustainable collaborations in the open data ecosystem The case-studies reveal the importance of institutions, for two purposes: (i) sustaining the collaborations identified in these case-studies, and (ii) replicating the types of collaborations and task distributions in these case-studies to other open data ecosystems. As mentioned above, for the IAD Framework, institutions encompass both formal and informal norms that define rules, shared understandings, and strategies guiding human behaviour and social decision-making. These shared rules and strategies are collectively developed, enforced, adjusted, and monitored to promote the sustainable use of information resources like open data. Consequently, collaborative actions within the 'action arena' can be distilled into practical institutional insights, potentially allowing for the replication of successful collaborations in different settings. First, the continued release of open datasets by public administrations is important, given that each case-study relied on open government data. These datasets can be released under legal regulations like the Open Data Directive. The scope of open datasets can also be increased voluntarily by public administrations, where additional datasets are made open as in the Breakthrough Open Data Project. Based on the importance of open government datasets for the collaborations studied in this report, legal regulations for the release of open data could be an important institutional factor. Continued implementation of laws such as the Open Data Directive together with the Implementing Regulation on High-Value Datasets combined with regular consultations with open data user and provider groups should be ensured. Further, provide new avenues by which public administrations can acquire data from non-government data holders and make them available to other users for certain types of re-use. There should be continued focus on strong implementation of such legal instruments as well. The combination of open datasets with other partially-open datasets can result in more value contributions, as evident in the beamm.brussels case-study. Second, the creation of coordinating entities in different sectors could boost collaboration. In the Breakthrough Open Data Project for instance, the Dutch government served as the coordinating entity that brought together public and private sector users of geodata and Dutch public administrations who are providers of open data and enabled the users to share their open data needs. Public administrations at local, regional, national and European level could proactively adopt such coordinating roles, to bring together open data providers and open data users in different domains to openly discuss and address the usersʼ needs. Going one step further, such an approach could also be adopted for the EU Implementing Act on High-value Datasets. A coordinating function could also be performed by public administrations, by providing funding to collaborative open data initiatives, as undertaken by the Brussels regional government in the beamm.brussels case-study. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 39 Third, the creation of partnerships between public administrations and NGOs could also boost collaboration to realise social value from open data. In the CityLAB Berlin case-study and in the Femicides in Europe case-study, there was a clear and formal partnership between the local government (in the former case) and the regional and European government (in the latter case) with NGOs, to work with communities and use open data to identify social issues of concern. In both case-studies, the projects received funding from the respective government bodies and therefore took the form of formal partnerships between the NGOs and the government bodies. These partnerships enable other actors, like non-specialised users, to contribute to open data initiatives. In many cases however, NGOs initiate projects and build partnerships with government bodies on their own, which could take the form of more informal partnerships. In such cases, there should be greater willingness on the part of government bodies to engage in such partnerships with NGOs. The partnership (whether formal or informal) should also include a feedback loop, where citizen-contributed data (as in the case of CityLAB Berlin) and NGO-created data (as in the case of Femicides in Europe) are integrated into open government data initiatives. Fourth, systems for data and information sharing serve as important infrastructures for collaboration. In three case studies (beamm.brussels CityLAB Berlin and Femicides in Europe), such systems were developed by an academic institution (in beamm.brussels), an NGO (in the CityLab Berlin case-study) and by coordinated action between journalists, NGOs and Eurostat (in the Femicides in Europe case-study). In terms of replicability, not all actors may have the necessary resources and skills necessary to build such infrastructures from scratch. As a result, within the European Union, shared infrastructures like the European Open Science Cloud, could also enable various actors to come together and use this infrastructure to analyse open datasets and generate knowledge from such analysis. Fifth and finally, a strong inter-organisational culture on the importance of realising economic as well as social value from open data in a collective manner is also important. For example, in the CityLAB Berlin case-study, CityLAB Berlin has a strong internal culture to create social impact. This motivated employees of CityLAB to propose, design and implement the project discussed in this report. Similarly, in the Femicides in Europe case-study, the project was led by journalists, who have a strong culture of using open data for data journalism. These types of inter-organisational cultures can be cultivated using formal institutions like legal frameworks, together with informal approaches to creating shared understandings among the various open data actors. D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 40 8 Conclusions 8.1 Objective of Task 5.1 Per the ODECO DoA, the objective of this Task 5.1 was to “design and review different models of allocating roles, tasks and resources in open data ecosystems.” In this report, we summarise 4 case-studies of collaboration between public administration and one other open data actor group for realising value out of open data. These case-studies serve as illustrative models for the allocation of roles, tasks and resources in an open data ecosystem. From these case-studies and by applying the theoretical concept of the ‘Institutional Analysis and Design Frameworkʼ, we propose some institutional factors enabling collaborations between open data actor groups, which could enable replication of these types of collaborations in other open data ecosystems. 8.2 Detailed conclusions An ecosystemic approach to open data recognises both the plurality of actors and plurality of motivations underpinning the generation and use of open data. Previous ODECO reports have identified different open data user needs. Previous ODECO reports have also proposed governance instruments to enable the generation of open data by non-government actors, as well as governance of open data in a manner that enables the realisation of both economic and social value from open data in a participative manner. In order to identify the different roles for actors to participate, we undertook a review of previous ODECO outputs on open data governance that identified roles, tasks and resources in open data ecosystems. We synthesised findings on the roles discharged by the identified open data actor groups in an open data ecosystem, and the types of value contributions made by these actor groups. Using a participative design approach, we developed and conducted a workshop at ODECO TW 5, Samos, to identify real-life examples of collaboration between open data actorsʼ groups to create value from open data. This workshop resulted in four case-studies. In each case-study, ESRs identified the actors involved, the tasks performed by these actors, the level of responsibility accepted by each actor, and the governance instruments that facilitated the distribution of tasks. From these case-studies and from a review of existing ODECO reports, we studied the different ways in which public administrations collaborate with other open data user groups (including companies, open data intermediaries, journalists, NGOs and citizens as non-specialised users), to realise social and economic value (and in one case, environmental value) out of open data in an open data ecosystem. These collaborations resulted in the release of more open data, improvement of the quality of existing open datasets, creation of visualisation and knowledge from open data, identification of social issues of concern, and the creation of civic communities. Finally, we synthesised some findings on institutional factors that enable sustainable collaborations in our selected open data ecosystem, which could allow our case-studies to become models of collaboration that could be replicated. Drawing from the Institutional Analysis and Design (IAD) framework, we defined ‘institutionsʼ as “widely understood rules, norms, or strategies that create incentives for behaviour in repetitive situations.” Institutions may be formally described in the form of a law, policy, or procedure, or they may emerge informally as norms, standard operating practices, or habits. The IAD theoretical framework has proved suitable to represent the relations between resources and communitiesʼ attributes, rules and patterns of interactions, particularly in the context of informational and knowledge resources (Frischmann, Madison, & Strandburg, 2014). D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 41 We extracted five institutional factors that we suggested could serve as starting points for the sustainable arrangement of actor roles, tasks and responsibilities in other open data ecosystems: 1. There should be continued focus on publication of open government datasets by public administration. To realise value out of open data, it is important that high-quality open datasets are easily available. Accordingly, public administrations should continue to release open datasets, based on legal mandates as well as on a voluntary basis. For instance, the Italian Military Geographic Institute represents one example of public administrations adopting a pro-active approach to release of open data. This public body creates and maintains a comprehensive geographic database for all of Italy called ‘Database di Sintesi Nationale (DBSN).2 This database integrates open government data from multiple public bodies, together with other open data such as data from OpenStreetMap and satellite data. As a result, the database contains data on a variety of parameters, such as public amenities, green areas, as well as data on types of property ownership. This database is machine-readable and released under an OdBL License. The database is used by the Italian Civil Protection Department during rescue operations, as it contains high-quality and updated geographic data. (DCPC 2024, pp 23; Santoro 2017). However, to ensure that public administrations prioritise release of open data, it may be important to evaluate different enforcement as well as incentive structures. 2. Coordinating functions should be discharged by public administrations to bring together open data providers and open data users to openly discuss their needs. Public administrations should routinely consult with open data users and providers from the citizenry, the private sector and the public sector to understand evolving needs of these stakeholders and undertake open data initiatives that respond to these needs. Public administrations should also provide funding to collaborative open data initiatives, which can then serve a coordinating function. For instance, the Glasgow Centre for Population Health coordinates civic projects involving public administrations, academic universities and civil society organisations. This enabled the Glasgow Centre for Population Health to create ‘Understanding Glasgowʼ3a website that hosts visualisations on health and life circumstances, encompassing visualisations on poverty, transport services, population and culture to name a few. (DCPC 2024, pp 31-32). At the same time, coordinating functions can also lead to open data ecosystems in unexpected ways. For instance, the UK mapping agency – Ordnance Survey – did create large geographic datasets, but many open data users were dissatisfied that these datasets were not freely distributed. This led to the birth of OpenStreetMap (see https://wiki.openstreetmap.org/wiki/History_of_OpenStreetMap) – an open platform for crowdsource geospatial data, which could be considered an example of a sustainable open data ecosystem. Future research can explore mechanisms for non-governmental actors to discharge coordinating functions. 2 https://igmi.org/en/dbsn-database-di-sintesi-nazionale/dbsn-database-di-sintesi-nazionale 3 https://www.understandingglasgow.com/ D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 48 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 49 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 50 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 51 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 52 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 53 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 54 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 55 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 56 D5.1 Models of allocating roles, tasks and responsibilities in open data ecosystems 57