Data governance at the local level: lessons from European cities

How to facilitate data sharing at the local level? What kind of organisations and rules should we put in place to ensure trust among the participants? How could we involve a broad range of stakeholders in defining these rules? Here are some questions that directly concern data governance.

At the micro-level, data governance is defined as “an organisation and a set of rules to manage data at every step of its lifecycle, from production to use and discard”. But, at the macro-level, data governance is also a political question, dealing with the power relations and (im)balance between different actors, including the public authority, the private sector and inhabitants of a territory. 

Users, citizens and inhabitants

“Data governance isn’t a new topic for Rennes”, explains Marion Glatron, smart city officer at the local authority. The city has been one of the pioneers of open data in France, launching its initiative in 2010. Since 2018, Rennes Metropole is experimenting with a local data public service. Still, Rudi raises a new series of questions. The first set deals with citizens participation. “We should probably distinguish between three different roles: inhabitant, consumer and citizen”, adds Marion Glatron. “An individual may contribute to the design of the portal as a potential user, he can have his say in the objectives and the rules of the portal as a citizen in its political dimension, and he can also activate its digital rights as an active user, a consumer of services”. Data portability is one of these digital rights. Rudi aims to give users the possibility to provide (and retract) consent to the use of their data and a clear and transparent view on the use of their data by third parties.

Raising awareness on data sharing is only the first step, but still a challenging one. “How can we build a community of non-experts around a subject - personal data sharing - perceived as complex and theoretical before we have anything tangible to show people?” explains Rennes Metropole. The launch of the prototype version of the platform in 2021 should help, as the presence among the partners of the NGO Fing (Fondation internet nouvelle génération), a key actor of self data in Europe.  

But Rudi isn’t only about citizens. The private sector actors’ participation is another challenge. The first step is to recognise the diversity of players, from startups, NGOs, SMEs, national and multinational companies. Some are entitled to a public service mission (energy, public transport, waste management); others are not. We need to better understand their motivations to contribute to the platform and their role as data producer and initiative partners. What are their incentives to share data? How to ensure local authority’s sovereignty on the local public data? How could the public sector access privately-held data for public interest purposes?

Marina Micheli gave clues to these questions. As a researcher at the Joint Research Center (European Commission), she studies data governance models. For this first Rudi Live Europe event, she shared her findings on the access to privately-held data by the public sector - aka B2G data sharing. “There are at least six operational models for B2G data sharing: data donorship, public procurement of data, data sharing pools, research data partnerships, challenges/hackathons and data sharing obligations in contracts”, explains Marina Micheli, adding that “these models are based on different collaboration dynamics between private sectors and local authorities”. 

For instance, in the UK, 4.3M Linkedin users profiles have been analysed to produce the London Tech Talent’s Economic Graph. By sharing, free of charge, these insights with the City of London’s policymakers, the platform intends to inform decisions about the labour market. All around Europe, mobile phone operators (Orange, Vodafone) are using anonymised location data to provide insights to cities. The Waze Connected Citizens program is a different operational model based on mutual data sharing between cities and the software company. 

Power relations between public and private sector actors

In 2019, Marina Micheli interviewed twelve European cities (including Amsterdam, Barcelona, Ghent and Rennes) to understand cities’ modalities of access to private sector data of public interest. Her study focused on the perspectives of local administrations and the power relations established between actors. “Reputation of a city is a key enabler for data donorship. Cities with a recognised advance in technology or participating in international collaborations are more likely to be beneficiaries of data donorship from private companies,” states the researcher. That also leads to an ethical dilemma for these privileged cities. Most of the time, less-known towns would have to pay to access the same data. One city representative expressed: “Do we really want to have free lunch if others are paying for it?”. Micheli’s research also sheds light on the power imbalance between big companies and local authorities when discussing data acquisition. Companies are sometimes reluctant to share extensive details about the completeness and quality of data during the negotiation phase. It makes it even more difficult for public authorities to understand the potential usefulness of a dataset and, hence, difficult to assess the real value for money of a deal.

Last but not least, some public authorities question the nature itself of the data produced or collected in the public space. As one city representative boldly expressed: “Data collected in the public space is from everyone, it’s not just from the company who happens to put a sensor”. Micheli concludes that if companies alone decide whether to donate/share their data (and with whom), the gap between “smart cities” and other cities might increase.

Olivier Thereaux is head of R&D at the Open Data Institute (ODI). ODI is a non-profit based in the UK working with governments, companies and civil society to develop an open and trustworthy data ecosystem. In the past few years, Olivier and his team researched how local governments collect, publish and use their data in London, Amsterdam, Hamburg, Copenhagen or Helsinki. “What we found is that there is a lack of collaboration, due to a segregation between the role of publisher and the role of the user”, explains Thereaux. “Furthermore, most access to privately-held data are based on coercion, such as public procurement rules, local legislation or any other kind of licensing rules”, citing the examples of Barcelona where the law gives city officials access to some Airbnb data or Washington DC’s dockless bike pilot which requires companies involved to have public APIs for available vehicle locations. The consequence is that private companies tend to see local authorities and their platforms as “parasitic”: “Companies think that their data is valuable and that cities are sucking their blood by stealing data or at least forcing them to give over data”. This perception of the relationship must change if we want to have a much healthier collaboration between public and private sector actors. Trust from individuals is the third challenge. In 2019, the ODI conducted a survey in France, Germany, Netherlands and the UK, asking people who they would trust with data about themselves. On average, 35% of the respondents trust their local governments: “that’s relatively high compared to other institutions such as central governments or retailers, but still the level of trust is quite low”. 

The Open Data Institute developed several tools to help change the current paradigm. The Data Spectrum is one of these. It presents a continuum of situations, from closed data to shared and open data. All data is on the spectrum. “We ask people: where on the spectrum should this data ideally sit, and how do we get it there? It’s a fairly powerful way of starting a conversation,” said Thereaux. 

Emerging best practices

The London Data Commission is taking a different approach, focusing first on the major challenges that the city faces such as housing, transport or skills shortage, and then discussing with private sector organisations producing data about these topics. Oliver Thereaux states that Covid-19 has been a tipping point: “perhaps because there is a sense of urgency or a clarity of purpose, I have seen more willingness and more actual data sharing practices than ever before”. He gave some examples: the Cuebiq x GovLab x ODI research on mobility data or the London Data Commission data sharing pilots.

“The main driver should not be to choose the right governance at the beginning. To some extent, that’s the mistake we made with open data portals, where governance - openness - was the point. Starting from the challenge and asking who holds the data might help us is much more productive than starting from the governance,” concludes Thereaux.

Useful links and resources

• Accessing privately held data: Public/private sector relations in twelve European cities, Marina Micheli, Data for Policy 2020 conference paper
• The Data Spectrum, Open Data Institute
• Who do we trust with our personal data? A survey by the Open Data Institute