What’s the best strategy for ensuring development projects adhere to ethical guidelines? This is a complex question as answers will vary by sector and context. Yet ethical issues such as bias, representativeness, and data availability and quality have direct impacts on development projects. Ethics and solid governance mechanisms that center community engagement must sit at the heart of these projects for them to be successful, sustainable and transformative.

In the seven years since Data-Pop Alliance’s work began, numerous learnings have led us to rethink our processes. As a result, most of our projects since 2020 include a Council for the Orientation of Development and Ethics (CODE). (See this 2020-2022 report for more information on CODEs.)

Building collaborative projects using the CODE 

A CODE is a group of independent stakeholders who voluntarily share their expertise in areas of direct relevance to a project. As an advisory group, it provides oversight to ensure a project abides by key ethical principles including fair and safe use of data and local context-specific concerns. CODE responsibilities vary slightly based on project needs but generally include the following: 

• Driving the creation, function and application of the legal conditions of project governance, including counseling on privacy-related issues such as access to and use of data, use of algorithms, biases, legal compliance, user authorization and community engagement.
• Improving the local relevance of the project including by suggesting ways to enhance project benefits and visibility and improve outreach to relevant groups.
• Recommending improvements to methodology and data sources and guiding development of a balanced and inclusive local data ecosystem.
• Advising on sensitive use cases and potential risks to intended target communities.
• Functioning as a forum for dispute resolution and community complaints regarding data use and engaging concerned stakeholders across sectors.
• Identifying strategies for a project’s future, including scaling, broader application and improvements to project design. 

Though the advisory role of the CODE is not legally binding, project staff give considerable weight to CODE members' feedback, integrating their recommendations whenever possible. As with any group of volunteers, CODEs will face challenges, from spotty meeting attendance to low levels of participation. Effective planning early in the project, preparation and active moderation are key ingredients to ‘healthy’ and impactful CODEs. 

CODEs are typically composed of 10 local experts representing academia, civil society, government and private enterprise. Data-Pop Alliance implemented CODEs last year in eight countries (shown in Figure 1). Most CODE meetings have been held virtually due to the COVID-19 pandemic. The CODE concept was born during the initial implementation of the Open Algorithms (OPAL) project, a data sharing initiative co-founded by Data-Pop Alliance that aims to unlock the potential of private sector data for public good in a privacy-preserving, participatory and sustainable manner.

Figure 1: Data-Pop Alliance implemented CODEs in Latin and South America, Sub-Saharan Africa and the Middle East in 2020.

The CODE in practice: Gender-based violence in Latin America

CODEs often steer projects in a new direction. We saw this with a recent project that aimed to identify personal and environmental factors associated with domestic violence reporting rates in three countries. This project required us to access data from government agencies providing services to survivors (police reports, hotline records, counseling and legal support centers, etc.) as well as from national surveys on family violence. Researchers used this data to map domestic violence reporting rates at the district level in three cities to analyze relationships between reporting rates and environmental risk factors.

Data-Pop Alliance set up three CODEs in Mexico City, Bogotá and São Paulo for the project, which was funded and supported by the GIZ Data Lab and the Unidas network between 2020 and early 2021. Involving the CODES early in the project resulted in a dramatic shift in our research focus and design that led to the production of meaningful and locally-relevant research while minimizing harm to individuals and communities.

Project leaders and CODE members established ethical guidelines for research and data management following best practices in the field including: (1) Non-maleficence or no-harm, (2) de-stigmatization, (3) confidentiality and privacy, (4) lawful, legitimate and fair use of and access to data, and (5) data quality and transparency. Consultation with the CODEs, however, quickly shifted our perspective on the ethics of the original project design. 

Our initial goal had been to create maps of domestic violence “hotspots” in the selected cities. Through periodic, virtual meetings with project CODEs, we realized that mapping geographical areas with higher rates of domestic violence posed a considerable risk of stigmatization to communities and individuals. Our team modified the project’s aim to avoid replicating long-standing discriminatory narratives about groups of people and geographical areas—a process that also incorporated suggestions from CODE members.

While we anticipated issues of confidentiality and privacy as primary ethical concerns, the principle of ‘no stigmatization’ became the key to ensuring this project did not violate other issues of no harm, confidentiality, privacy and more. As a result, the project produced an analytical model with insights into individual and contextual factors that may hinder or enhance reporting and registering domestic violence incidents. These insights in three countries also led the project to outline policy recommendations to improve the quality of gender data. Reports on this project will be published later this year.

A ‘North Star’ in project design and implementation

CODEs can serve as North Stars and sounding boards. Their members apply a fresh set of eyes to a project’s design, development and implementation and periodically demand that the implementing team question their actions, decisions and findings. The best performance of a CODE relies on its members being independent with low stakes in project outcomes. As critical outside voices, they provide important feedback that enables the project to succeed. In sum, CODEs ensure that data in development projects adhere to ethical standards by fostering critical self-assessment and accountability through a participatory approach. 

“The CODE concept and operations manifest the values of the UN in the digital age, building trust in data and partnerships.” 

- Fouad Mrad, Senior Programme Manager United Nations Economic and Social Commission for Western Asia

Ivette Yanez is Project and Communications Manager and Emmanuel Letouzé is Director and Co-Founder of Data-Pop Alliance. Follow Ivette on Twitter at @ivetteyanez_ and Emmanuel at @ManuLetouze.

This post was first published by the Open Data Institute on March 23, 2021 on its blog here. 

Data institutions are organisations that steward data on behalf of others, often towards public, educational or charitable aims.

In practice, data institutions steward data in different ways, including:

• Protecting sensitive data and granting access under restricted conditions.
• Combining or linking data from multiple sources, and providing insights and other services back to those that have contributed data.
• Creating open datasets that anyone can access, use and share to further a particular mission or cause.
• Acting as a gatekeeper for data held by other organisations.
• Developing and maintaining identifiers, standards and other infrastructure for a sector or field, such as by registering identifiers or publishing open standards.
• Enabling people to take a more active role in stewarding data about themselves and their communities.

Although the term is relatively new, many data institutions exist across the private, public and third sectors. Organisations like national mapping agencies, statistics agencies and archives are perhaps our oldest and most well-known data institutions – in some cases they’ve played these roles on behalf of the public for hundreds of years.

Data institutions come in many shapes and sizes, and each will have had a unique journey and faced different challenges along the way. Some, like Open Banking Ltd, have been brought about by government regulation and funded by industry. Others, like Salus Coop, are small data institutions still trying to build a critical mass of users and attract investment. 

Here we document the genesis of a variety of data institutions, which we’ve chosen to illustrate the different paths they can take and some of the different ways in which they are funded. We hope that by telling these origin stories, we will increase understanding of how data institutions emerge, and inspire people to think about starting their own data institutions in the future.

UK Biobank

UK Biobank is a registered charity founded in 2006. It aims to improve the prevention, diagnosis and treatment of a wide range of serious and life-threatening illnesses.

From across the UK, 500,000 people aged between 40 and 69 years old underwent testing, provided blood, urine and saliva samples for future analysis, gave detailed information about themselves, and agreed to have their health monitored to create the database that UK Biobank stewards.

Since 2012, researchers have been able to apply to use the UK Biobank database. Studies that use the database typically compare a sample of participants who developed a particular disease with a sample who did not, in order to measure the risks and benefits of certain interactions with genes, lifestyles and medications.

UK Biobank is funded primarily by the Medical Research Council and the Wellcome Trust, with a total funding amount of £244.3m over its lifetime. It supplements this long-term investment with project funding to deliver specific enhancements to the database and uses a cost-recovery model to charge users for access to the data, with reduced fees for student projects and those from low- and middle-income countries (LMICs). 

UK Biobank’s evolution demonstrates the value in providing significant, long-term investment from government and philanthropy. According to UK Biobank, over 670 international research groups have accessed health records during the pandemic alone, leading to more than 60 scientific papers being published in the public domain.

HiLo Maritime Risk Management

HiLo is a startup that supports the sharing of safety and accident data in the maritime sector.

Founded in 2014 as a small-scale joint project by Shell Shipping, Maritime Maersk Tankers A/S, and Lloyd’s Register Consulting, it became an independent company in 2018.

Prior to the formation of HiLo, shipping companies were only able to draw on their own operational data to try to reduce the frequency of accidents on their ships. By pooling data from across these companies and analysing the aggregated dataset, HiLo is able to provide meaningful and actionable benchmarks, and other insights, to its member companies. The more organisations that share data, the better and more reliable the insights they receive.

Around 55 companies are members committed to sharing data. The data is shared by these contributors directly to HiLo once a month, a manual process via the HiLo portal or automatically via an API. The data is then processed and analysed using HiLo’s risk model algorithm. Financially, HiLo receives subscription fees for its insight service, but hopes to diversify its revenue streams to include a more balanced mix combining subscriptions with revenue from new products and services generated by the data it brings together.

HiLo has helped to save lives and money – so far, it has reduced lifeboat accidents by 72%, engine room fires by 65% and bunker spills by 25%. It shows how an important, impactful data institution can be built from small beginnings as an industry project. 

Open Banking Ltd.

Open banking represents a trend towards consumers gaining more control over the data generated by their current accounts and other financial products.

In 2016, the UK’s Competition and Markets Authority (CMA) established Open Banking Ltd, to deliver on its vision of helping people save, borrow, lend and invest money securely while improving efficiency, increasing competition, and stimulating innovation within the sector.

Open Banking Ltd is a private company, governed by the CMA and funded by the UK’s nine largest banks and building societies. It is responsible for developing and maintaining the data infrastructure – including technical standards and guidelines – that enables data to flow between the banks and other organisations. It makes it easy and safe for individuals and small to medium sized enterprises (SMEs) to share the financial information held by their banks with third-party services. As well as funding Open Banking Ltd, the banks are also mandated to comply with its guidelines. 

The data infrastructure maintained by Open Banking Ltd has supported the industry to build useful new applications for customers. One example is Account Information Service Providers (AISPs), such as an app from fintech startup Bud. This app was designed to allow customers (who opt in via their bank’s platform) to see a dashboard showing the status of multiple accounts from different bank providers, and manage their finances more efficiently. Other services using Open Banking data help small businesses access loans at better rates, or recommend accounts based on spending patterns.

Open Banking Ltd demonstrates that governments have a role to play in mandating the creation and funding of data institutions where data ecosystems are not working effectively. 

The UK Office for National Statistics

The Office for National Statistics (ONS) is the UK’s largest producer of official statistics, responsible for collecting and publishing data related to the economy, population and society.

The ONS was formed in 1996 following the merger of the Central Statistical Office (CSO) and the Office of Population Censuses and Surveys (OPCS). It functions as the executive office of the UK Statistics Authority, that reports directly to parliament following the Statistics and Registration Service Act 2007.

Examples of the data collected and used by the ONS include information from the decennial population census (a census taken every 10 years), data from businesses, government departments, public sector bodies, such as the NHS, and the registers of births, marriages and deaths. This data is often central to debates about allocation of national resources and economic decisions. It is used to measure changes in society such as migration, employment or illness rates and longevity. ONS data is also employed in epidemiologic studies – for example it has published extensively about the coronavirus pandemic, tracking everything from infection and death rates to social and economic impacts.

The ONS produces and publishes a wide range of data. It publishes its own statistics and reports, and also makes published data available for other users via the Secure Research Service (SRS). The SRS makes anonymised, unpublished data available to accredited researchers engaging in research projects for the public good. The ONS also has a Data Science Campus, which oversees a series of data projects that provide insight into key policy themes, and a joint team with the Foreign, Commonwealth and Development Office, which aims to use data for global public good.

The ONS example highlights the fact that some of our most important data institutions are publicly owned and governed. In thinking about the need for new data institutions we should not forget the role of the state as a steward of data on behalf of others.

Open StreetMap

Launched in 2004, OpenStreetMap is a collaborative project that aims to create a free, editable map of the world. 

OpenStreetMap was started in the UK and now has over two million registered volunteers worldwide, with over 5,000 users a day, making it the world’s largest crowdsourced open database. These independent volunteers collect geospatial data from scratch, performing ground surveys with tools such as handheld GPS devices, cameras, and notebooks. The map and database they collectively build is made freely available under an open licence.

As the project has grown, commercial and government organisations – including Yahoo! and the Federal Government of the United States – have made the data they hold available for manual editing and automated imports via OpenStreetMap. OpenStreetMap data is widely used, including by high-profile organisations such as Facebook, Apple, Microsoft, Amazon and Uber, both as a map and as a data source for visualisation, research and analysis.

The project is supported by the OpenStreetMap Foundation, a not-for-profit company registered in England and Wales. The foundation has members from around the world and an elected board of directors, and is both the legal entity for OpenStreetMap and the custodian for the computer services and servers which host it. It also provides a vehicle for fundraising and donations to support the project, organises an annual conference, and supports various working groups looking after communications, licensing and other functions.

OpenStreetMap is a fantastic example of how a community can come together to create a data institution, and how a data institution can grow internationally, by collaboratively sharing and maintaining data.

Salus Coop

Salus Coop is a non-profit citizen data cooperative for health data, founded in Barcelona in September 2017. It set out to create a citizen-driven model of collaborative governance for and management of health data ‘to legitimise citizens’ rights to control their own health records, while facilitating data sharing to accelerate research innovation in healthcare’.

Salus Coop has developed a ‘common good data licence for health research’, which applies to the data that members donate, and specifies the conditions that any research projects seeking to use the member data must adhere to. These include only using data for research on a non-commercial basis, making all results accessible at no cost, anonymising data before use, and allowing members the rights to cancel or change the conditions of access to the data about them at any time. 

Salus Coop has been supported by the Mobile World Capital Barcelona Foundation and Ideas for Change, and also invites health authorities to collaborate with its efforts to support people to participate by sharing data about themselves. In 2020 it started the CO3 (Cooperative COVID Cohort) project to convene a group of citizen data donors and create a new resource to use for research into Covid-19. 

Salus Coop is an example of a new breed of data institution emerging – sometimes referred to as ‘data cooperatives’, ‘data unions’ and ‘bottom-up data trusts’ – that enable people to take a more active role in stewarding data about themselves and their communities. Many of these efforts are still trying to reach a critical mass of users and become sustainable. 

• Alex Vryzakis is ODI Associate-Communication while Jack Hardinges is the Programme Lead for Data Institutions at ODI. Find out more about the Open Data Institutes ongoing work on data institutions here. 

Big Data promises to improve government service delivery, complement official statistics and facilitate development in a range of sectors. Countries around the world stand to benefit from ‘Big Data for development’ in health, food security, infrastructure, transportation, humanitarian response and more. Yet, interest until now in capturing and analysing Big Data for public purposes has largely been driven by Northern institutions. A two year-project, Harnessing Big Data to meet the Sustainable Development Goals - Building Capacity in the Global South, funded by the International Development Research Centre (IDRC) and led by the African Institute for Mathematical Sciences (AIMS) and four other institutions from the Global South (CEPEI, LIRNEasia, LDRI and the centre for internet and society), sought to address this challenge by developing capacity among researchers in the global South focusing on the use of Big Data to inform progress in meeting the Sustainable Development Goals.

AIMS is a post-graduate education network and research institute with five centres in South Africa, Senegal, Ghana, Cameroon and Rwanda that served as the regional hub for this project in Africa and aligned the focus of its work to the African Development Bank’s High-5 Priorities. AIMS worked with partners to increase the number of Big Data scientists on the continent, catalyse research on Big Data to address policy, infrastructure, funding and human capital gaps and provide a platform for practitioners to network, learn and coordinate activities. 

In this interview, AIMS director of Industry Initiatives Charles Lebon Mberi Limpolo shared some key lessons from this project on the drivers of better up-take of Big Data in Africa. 

You recently completed the Harnessing Big Data to meet the Sustainable Development Goals project. Can you tell us more about this project’s main goals and aspirations?

The impulse for this project came from the realization that voices from the Global South were not sufficiently recognized in discussions around Big Data at the global level. AIMS came together with partner institutions from Colombia, India, Sri Lanka and Kenya to challenge this status quo. 

We had many objectives and aspirations at the start of the project—notably in terms of prioritizing gender in research, engagement of the public and private sectors, influencing national, regional and global policy processes and applications of Big Data techniquest to local development problems. One of our main objectives was also to strengthen the capacity of local actors to use Big Data. Each partner could tailor these objectives to the needs of their specific region. For the African Hub of the project, AIMS focused on developing capacity (through training and other capacity-building activities) and on supporting policy and innovation for development and key research initiatives.

What are the biggest achievements of this project, and what are you most proud of? 

We delivered a short training course (Big Data Analytics with Python) and created a Big Data Education executive programme for C-level executives. We ran an innovation programme contest to tackle four development challenges focused on financial inclusion, food security and youth employment. Ten groups of young people and data scientists submitted proposed solutions to these challenges using Big Data. The two winners were included in the Open Innovation Program - Make-IT Africa, sponsored by GIZ Rwanda.

Within and outside of the context of the innovation programme, we built close relationships with private sector agencies that shared their data and engaged in project activities. Our private sector partners included financial institutions, Mobile Network Operators and media companies from several African countries. In many cases, we convinced institutions to share their data for the first time and even to fund the innovation programme. This was the case, for instance, for Orange and the Bank of Kigali. We also published the first Pan-African report on the State of Big Data for Development in Africa. 

It is difficult to pick what I am most proud of. Working with young data scientists in the innovation programme and coaching them to develop solutions to key development challenges was exciting and important. The winners were given access to finance, market and skills resources and their solutions are now registered companies in Rwanda and part of the national start-up ecosystem. There was a big sense of excitement in the room when they presented their solutions. The impact of developing the training courses was also rewarding. These trainings were key to filling the needs of and increasing demand from the younger generation. 

What are the key lessons learned from the project in terms of Africa’s unique perspective on Big Data for development? 

There are two lessons learned which are unique to Africa. One concerns the importance of training and working with C-level executives to increase up-take of Big Data solutions. Capacity development is not only about technical skills: Leaders can boost demand for technical skills and promote development of new services and products. In other areas of the world such as India or Southeast Asia, leaders seem to have a better understanding of the potential of technology. Africa, in comparison, presents a bigger challenge in converting C-level executives.

We also learned that there is growing demand for courses among African youth that focus on Big Data. However, the average person has very limited access to such courses. Obstacles include the limited number of available trainings, language barriers and infrastructure barriers that limit the reach of online training. That’s why increasing access to this type of education remains a priority.

What are the main barriers to Big Data for development that this project highlighted in the African context?

There are two main categories of barriers: technical and non-technical. Technical barriers include a lack of technical and hard skills as well as insufficient equipment and poor technological infrastructure. Non-technical barriers encompass a broad range of issues such as the lack of data culture and data awareness (especially among leaders), few strategic partnerships to promote Big Data up-take and limited policy frameworks to support data-sharing. Limited funding for research and innovation also remains a key challenge.

What do you believe should be the role of the private sector in the Big Data 4 Development domain? 

The private sector has an important role to play in fostering the use of Big Data in Africa. First, companies must participate in discussions around development priorities and objectives. Development must be seen from all angles, and private sector objectives are not always incompatible with broader socio-economic development. Secondly, the private sector contributes to the demand for skills to process, analyse and use Big Data that can drive efforts to upskill the population. This is why we need to educate C-level executives on the potential of new technologies. Furthermore, private sector stakeholders should participate in the policy debate and help develop data-sharing frameworks that are fit for purpose.

What needs to happen next for Big Data for decision-making to be rooted in the public and private sectors and in development practices in Africa? 

We need to ensure funding is available to replicate some of the good experiences which emerged in the last few years, such as our innovation programme. We also need to spend more time discussing and enabling more data-sharing in the private sector. This requires finding new types of partnerships and frameworks to embed security and privacy concerns. Ultimately, we need to educate people to the language and requirements of Big Data. We are no longer speaking about data as “statistics” anymore. Big Data consists of large quantities of information from different sources, and we must ensure such data can be used within the context of solid policy frameworks and without compromising individual privacy.

• Dr. Charles Lebon Mberi Kimpolo is the Director of the Industry Initiative at the Global Network Secretariat of the African Institute for Mathematical Sciences - Next Einstein Initiative (AIMS-NEI) and a member of the Data Values Project Technical Advisory Group. 

Data should be governed with transparency, inclusivity and accountability due to growing concerns over data protection and digital rights—particularly during the COVID-19 pandemic. Here are four approaches to aligning data values across communities of practice that should be on policymakers’ radars. These draw from a recent Data Values Project webinar, hosted by the Restore Data Rights movement. 

1. Manage expectation gaps. As more players enter the space of data production and digitization, expectations on data governance and sharing will differ. There will likely be conflicting interests in the supply and demand of data—especially around sensitive data and what information can and should be made publicly available. Think, for instance, about the controversies surrounding whether mobile phone-derived data should be used for COVID-19 contact tracing. Ensuring that these expectations are transparently discussed and creating space for inclusively resolving tensions with adequate accountability mechanisms for when things go wrong at the outset reduce suspicions about how and why organizations are collecting and using data. Informing citizens on how their data will be used helps build trust, but we know that informing citizens and even getting informed consent is not enough. Data collectors and controllers must take on greater responsibility to safeguard citizens’ rights through mechanisms that create transparency and promote inclusion, for example, by involving trusted intermediary organizations working at the local level that can help to engage with and solicit input from target communities. 

"Productive collaboration [on data governance] means meeting the needs of governments, being confident to say no to sharing raw data, and respecting the privacy of individuals."  -Sema Sgaier, Surgo Ventures

2. Get the basics right. In Kenya’s journey, establishing systems for data governance meant starting from scratch, and this took time and effort. The Data Protection Act was enacted in 2019, but it took more than a year for an officeholder to be appointed amidst the COVID-19 pandemic. Landmark cases underlined the lack of trust and divided opinions around data protection in Kenya. All these made the process even more difficult and required deliberate efforts to raise awareness: Kenya's government developed manuals, regulations and guidelines and consulted with citizens as part of this process. 

"Data protection is a novel concept for Kenya. Sometimes even public authorities seek guidance on the application of the Data Protection Act." - Rose Mosero, Government of Kenya

For countries with no previous data governance systems, building is also a balancing act between the rights of data subject sand public health amidst the pandemic, between rights and needs of data users and producers and between digitization and data-driven governance efforts and safeguarding data rights. 

3. Evolve statistical systems to retain trust. Since the start of Agenda 2030, there has been unprecedented pressure to deliver more and better statistics and to expand statistical ecosystems to new stakeholders.

The statistical community faces four pressure points including greater demand for granularity in data and for placing data into context and drawing out imprtant patterns (i.e. impacts on the environment, infrastructure, mobility, etc.). Third, there's heightened demand for timely data to enable predictions. And, finally, the number and types of data users and producers have expanded. Along with these pressures, there is additional demand on statistics offices to remain trusted sources of data by putting into place appropriate legal systems to respond. 

"Not many countries in Africa have statistical Acts that suit the digital contexts. If NSOs do not have the right legal systems, it is difficult to protect confidentiality." - Molla Hunegnaw Asmare, UNECA

Sustaining trust in statistics has three dimensions: trust in the quality of data that underpins statistical production, in the integrity of statistics and in the presentation and interpretation of data. Trust also requires safeguards for protecting personal data, investments in data literacy and regular dialogues and communication with citizens.

4. Transition from data governance elitism to a bottom-up approach. The real value of data governance is realized through ensuring that citizens are included in and become part of these conversations and have a say in decisions about their data and digital rights. The data for development community has often been strong advocates of the idealistic view that increasing data collection and usage will lead to social good. But this cannot be achieved at the expense of people and their rights. The value of engaging citizens in establishing core data values and principles and bridging the gap between data optimists and data pessimists is that we can better articulate how to move forward in way that promotes transparency, inclusion and accountability into an advocacy that is more than the sum of its parts. 

These four approaches to align data values across communities of practice demonstrate ways that individuals, institutions and governments can build transparent, inclusive, and accountable data systems and values at a time in which societies are being rapidly shaped by data. We would love to hear from you! Please continue the discussion using #DataValues #RestoreDataRights on social media or by contacting us at [email protected].

• This post is a summary of discussions from a June 22, 2021 webinar hosted by Restore Data Rights (#RDR), a grassroots movement for data rights in Africa post-COVID, and by the Global Partnership for Sustainable Development Data (GPSDD). The webinar brought together stakeholders from across the digital rights and data for development communities in Africa and beyond to discuss data governance approaches that balance digital rights with innovation and opportunities for future actions. Watch the full webinar on Youtube here. Speakers included Sema Sgaier of Surgo Ventures, Molla Hunegnaw of UNECA, Rose Mosero from the Office of the Data Protection Commissioner, Government of Kenya and Tom Orrell of DataReady, Al Kags of the Open Institute, ‘Gbenga Sesan of Paradigm Initiative and GPSDD secretariat team members Claire Melamed, Jenna Slotin and Davis Adieno.

This is the first in a series on data-sharing between public and private sectors focusing on emerging approaches and uncovering key lessons for regions and stakeholders around the world.  

The key questions surrounding business-to-government data-sharing in Europe

It seems there’s no respite for the data community these days. The COVID-19 pandemic has put data in the spotlight as a key ingredient for effective public response amidst rising concerns around personal data access, protection and privacy. The public debate has rarely been so lively nor data practitioners so busy. 

The European Commission’s recent launch of public consultations on the Data Act adds another heated topic to the current debates, with the possibility—among other provisions—of imposing binding rules to give governments access to private-sector data. The Data Act is a legislative proposal by the European Commission that “aims to create a fair data economy by ensuring access to and use of data for legitimate purposes, including in business-to-business and business-to-government [B2G] situations.” Building on the work of the Expert Group on B2G Data Sharing  and its 2020 report, the public consultation articulates some of the fundamental questions around government access to business and privately-held data, including:

1. In which cases or areas should B2G data-sharing be compulsory? The consultation establishes a long list of possible situations, areas and contexts including: (1) emergencies and crisis management, prevention and resilience, (2) production of official statistics, (3) protecting the environment, (4) promoting a healthier society, (5) providing better public education services, (6) fostering an inclusive society and (7) evidence-based public service delivery and policymaking. This list of use cases requires accessing data from a range of private sector players such as Mobile Network Operators for mobility data, retailers for consumer statistics, manufacturers for emissions data, companies for employment data, etc. 
2. Should the public sector have to pay to access privately-held data? And if so, how much?  The proposal includes a range of cost model options to compensate the private sector for providing data to the public sector: (a) for free, (b) at market prices, (c) at marginal cost (corresponding to the cost of production and a cost-recovery model for the private sector), (d) at a preferential rate at or below market price or (e) at differing costs (including at no cost) depending on circumstances. 
3. Which safeguards should protect the private sector when sharing data? Safeguards might invoke special rules for commercially sensitive information, set criteria for proportionality and reasonableness of data-sharing requests, oblige public authorities to report transparently on the use of data, and limit public sector data usage and data sharing, among other options.  

To European legislators, the notion of “clear public interest” is the cornerstone of B2G data-sharing and the justification for imposing obligations on businesses. This idea comes from the work of the Expert Group on B2G Data Sharing. However, the expert group has highlighted that, “while ‘public interest’ broadly refers to the welfare of individuals in society, its exact boundaries remain largely undefined, being heavily dependent on socioeconomic, cultural and historical factors.” Furthermore, as noted in recent legal research, “governments are free to define public interest and designate public tasks accordingly.” This is why it is extremely difficult to identify which use cases justify obliging businesses to share data. Balancing private and public interests in a rule of law context is the biggest challenge to establishing this legislation in Europe. 

Takeaways for other countries

That the European Union is considering obliging businesses to share data with the public sector through legislation provides food for thought to other countries struggling to gain access to privately-held data. Paradoxically, it would be easier to adopt similar legislative measures in countries where the rule of law is weaker and where—in the relationship between public and private stakeholders—the public sector has the upper hand. 

Because this example could lead other nations to adopt similar policies, it’s important to debate and learn from the European experience. The public consultation will be open until September, and the European Commission’s analysis of the responses and concrete policy options will follow later in the year. Nonetheless, it's not premature to identify a few takeaways and considerations of potential help to countries weighing similar decisions:

1. Public vs. private interest trade-offs: As the notion of ‘public interest’ is the starting point for B2G regulatory initiative in Europe, its meaning will vary a lot across countries, situations and cultures. If not balanced against private interests, the concept of public interest can lead to regulations encompassing the vast majority of private-sector data. It is important for governments to draw a line between 'must have' and  'nice to have' datasets from a public interest perspective. Such definitions should include objective proportionality tests to determine whether data requests are valid. Civil society and non-governmental organizations can help governments to establish such boundaries and should be included in debates on the balance between public and private interests. 
2. Compensation for private companies: Compensation in exchange for access to private-sector datasets is not taboo. Reality is complex: While some situations might necessitate the free provision of data (i.e. for emergency response in the context of data for development) other use cases may call for access at marginal cost or even at market price. The level of compensation depends on the relevance of the datasets to the public interest, the public sector’s available resources, the urgency of the situation and the time horizon of the data-sharing initiative. For instance, one-off data-sharing for emergency response might be appropriate on a pro bono basis, but sustainable and regular data-sharing for the production of statistics might require a cost-neutral approach for businesses. 
3. Safeguards to prevent abuse: Safeguards must ensure that the public sector does not abuse its privileges of access to business data. Such safeguards are important not only for businesses but also for citizens and consumers. People may entrust their personal mobile data to telecommunications operators, but not—as the COVID-19 crisis has shown—to the public sector for a variety of reasons. Public sector use of private data should therefore be framed as a set of rules to allow both business and private citizens to appeal and defend their interests. 

These takeaways provide initial food for thought to non-European Union countries and development practitioners working on public-private data-sharing agreements. The outcome of the public consultation and the release of the text of the Data Act will shed further light on where European legislators will land in imposing obligations on the private sector. 

A balance in Europe between public and private interests may not fit the needs of low and middle-income countries. Nonetheless, the key concepts and guiding questions posed in the European debate will foster dialogue across regions and may lead to the emergence of a new General Data Protection Regulation situation in which a number of countries follow the European Union’s lead. 

This post originally appeared June 17, 2021 on Medium. 

Early in my career, a mentor said to me, We measure what we treasure, and we treasure what we measure.

It was a simple play on words and a chicken-or-egg sort of statement, but the message resonated. At that moment, early in the Millennium Development Goals-era, we realized that simply counting girls in and out of school was nowhere near enough. We needed to understand the whole child. Therefore, gender mattered, but there was so much more: Was she a teenage girl from a village who spoke a language other than the national language? Did she have special needs that required actions to facilitate learning? Was she actually learning? Was she a citizen? Had her mother finished school? Did her family live above or below the poverty level? And, and, and... We needed measures—well beyond the superficial—for what we treasure.  

At the heart of ‘we treasure what we measure’ is a case for doubling and tripling our efforts to ensure no one is left behind in development data. This compels us to seek data and consider all dimensions and combinations of dimensions of the human condition to ensure that realities of everyone are brought into the light through data and treasured. And in doing so, perhaps there is hope for inclusion and equity in the realization of the broad goals for sustainable development.

• Jon Kapp is Executive Director of Community Systems Foundation and a member of the Global Partnership's Technical Advisory Group which leads the Data Values Project. 

On 22 June 2021, I organised an event entitled Dissecting Digital Power Inequities: Reflections from Digital Rights Experts for Development Practitioners as part of my engagement with the Data Values Project. This was an event I’d wanted to organise for a while and was quite personal to me. While I’d encourage anyone reading this piece to watch or listen to the whole event by following the hyperlink above, for those who may not have the time to watch a 90-minute event, I wanted to take a few moments to share why this conversation was so important to me. I hope that this piece provides colleagues and friends in my professional circles with some understanding of why I believe so strongly that human rights need to be at the heart of the Data Revolution for Sustainable Development.

When the UN Secretary-General called for a Data Revolution for Sustainable Development in 2014, it feels to me that the world was a very different place to what it is now. Within the sustainable development sector, an optimistic atmosphere prevailed as the nations of the world gathered at the UN General Assembly in 2015 to approve a science-based, ambitious universal Agenda for Sustainable Development – the 2030 Agenda.

At around the same time in my personal life, I made my first visit to Palestine, where I have family. To get to the West Bank, I traveled through Tel Aviv in Israel. Upon arrival, I was interrogated by Israeli security for around four hours. They wanted to know who I was and who my relatives were. I was shown pictures of family members, told telephone numbers, addresses and other information. The Israeli state clearly knew more about my family than I did as I’d not met any of my relatives before. In addition to personal questions about my own life, I was also asked questions about my parents – where they lived, where and how they met, what they did and the like. Once the Israeli state had worked out how I fit into the profile they had of my family and were satisfied that I did not pose a risk to security, I was allowed to enter.

I didn’t really think about this experience again for a couple of years. However, shortly after Donald Trump became the 45th President of the United States in 2017, I found myself thinking about it again. Shortly after the so-called ‘Muslim Ban’ was brought into force, I found myself organising a work trip to New York. I’d heard horror stories from Arab American friends of mine about how the travel ban was being implemented – it focused less on religious beliefs and more on ethnic background and nationality. Before departing, I wrote up a contingency plan for my wife including the names and numbers of lawyers to call should I be detained. As I packed, I made sure I had a change of clothes and toiletries packed in my cabin luggage in case I had to spend a night in an airport or holding cell. As the airplane sat on the tarmac preparing for take-off, I deleted text messages, emails and social media chats. I then deleted social media apps.

Why was I so afraid? I’m not a criminal. I’d done nothing wrong. I was afraid because I remembered all the information that Israel held about me. I was afraid because I instinctively suspected that Israel would have shared that information about me with US authorities. As it turned out, my fears were not entirely justified and I wasn’t detained or interrogated. I did however spend four years being subjected to the opaque Secondary Security Screening Selection (SSSS) process every time I travelled to the US, meaning that I had to go through enhanced security checks.

Finding myself in this situation - of being afraid for my physical safety at the hands of the world’s superpower for no reason other than my ancestry - was a very uncomfortable feeling. It also changed the way I perceived digital data about myself. I no longer viewed it as something innocuous that floated around in cyberspace. I started to see it for what it is – a representation of how others choose to classify me based on assumptions that they have drawn from my physical characteristics, ethnic and cultural heritage, political views and opinions, and other ‘data points’. This data holds real power. It holds power that can easily be turned against me if outside of my control - in the hands of powerful people whose inferences derived from my ancestry are based on their own biased, racist assumptions for instance.

“I started to see it for what it is – a representation of how others choose to classify me based on assumptions that they have drawn from my physical characteristics, ethnic and cultural heritage, political views and opinions, and other ‘data points’. This data holds real power.”

I recognise my privilege in all of this. I came to no real harm, although I did feel great pressure to self-censure details of my heritage in the Trump years – for instance making sure that my beard looked more ‘hipster’ than ‘Arab’ when traveling. Many aren’t as fortunate or privileged as I. Within my professional life, I can think of numerous examples where people face far greater disenfranchisement, fear and loss of rights all enabled by ‘data’:

What if I were a Muslim resident in the Indian state of Assam or Indian-administered Kashmir, who has had my citizenship stripped from me under the pretext of the Aadhar digital identity system being rolled out.

What if I were a member of the Nubian minority in Kenya, also afraid that the roll-out of a national ID system there might further disenfranchise me and render me invisible.

What if I were an African American challenging the use of racist facial-recognition algorithms used by the police and justice system in America.

What if I were a much persecuted and extremely vulnerable Rohingya refugee in Bangladesh, who, traumatised, destitute and desperate, entrusted the UN Refugee Agency (UNHCR) with my most personal biometric data only to later discover that it had been handed over to my torturers and tormentors in Myanmar.

What of these people? As development professionals, do we not owe them a duty to at least consider the risk of harm and threats that they face when we work with governments and private companies to develop data and digital infrastructure, ostensibly for sustainable development?

In the past eighteen months or so, the COVID-19 pandemic has turbocharged the blurring of the lines between cyberspace and the physical world. In many parts of the world, whole populations are being tracked digitally to ensure that they abide by lockdowns, curfews and the like. Examples of abuse of power abound. All of this surveillance is now digitally enabled.

What does all this mean for the community of dedicated, optimistic, science-driven organisations and individuals who make up the Data Revolution for Sustainable Development, of which I am a part? I think it means that we have to make the moral case for the Data Revolution to have human rights at its heart a lot more powerfully. We need to listen to, and ally with, digital rights activists who know the risks. We need to rely less on ‘quick wins’ and focus on designing data infrastructures that are transparent, participatory, inclusive and accountable from the outset. We need to make sure that our interventions do not propagate harmful, extractivist digital economy business models that undermine trust in digital data and cause enormous damage to the cause of sustainable development.

“We have to make the moral case for the Data Revolution to have human rights at its heart a lot more powerfully. We need to listen to, and ally with, digital rights activists who know the risks. We need to rely less on ‘quick wins’ and focus on designing data infrastructures that are transparent, participatory, inclusive and accountable from the outset."

I’m delighted that I’ve been able to be a part of taking the first step in this direction by organising Dissecting Digital Power Inequities: Reflections from Digital Rights Experts for Development Practitioners. I am grateful to the amazing activists and thinkers who volunteered to share their knowledge and experience with my community of practice. I’m grateful to the Global Partnership for Sustainable Development Data for providing a space for this discussion to happen. I hope that it is the first of many and that the outcomes of the Data Values Project are not just performative, but contribute to a change in how we do business in the data revolution.

• Tom Orrell is Managing Director of DataReady and co-leads the Data Values Project track on ‘Data that is well-governed.'

“Data are the lifeblood of decision-making and the raw material for accountability. Without high-quality data providing the right information on the right things at the right time; designing, monitoring and evaluating effective policies becomes almost impossible.” 

  - A World That Counts, 2014

These words still resonate almost seven years after the UN’s landmark report, A World That Counts. Harnessing the data revolution for sustainable development has always put using data at the center. Yet, in practice, the development community has placed much greater emphasis on the early steps in the data value chain, focusing more on collection and publication of data than its usage. This is understandable. After all, patchy, unreliable, and inaccessible data can’t be used effectively. But the last few years have demonstrated that access to high-quality data does not guarantee its use. 

Why access to good data doesn’t guarantee better development outcomes

Despite increasing amounts of available data, data-driven decision making is still not the norm for many organizations. More often, use of data by organizations remains scattered and doesn’t translate into evidence-based policies or decisions. 

Studies on barriers to effective data use generally distinguish a few broad categories: 

• Barriers linked to the data themselves - data accessibility, quality and appropriateness.
• Barriers related to stakeholders’ skills, capacity and incentives to use data, as who is using the data also matters. 
• Barriers concerning institutional and organizational resistance or lack of data culture.

Other barriers include: political economy considerations; a lack of funding; infrastructural barriers; or situations of conflicts within countries. 

Among low and middle income countries in particular, significant barriers exist due to limited access to data and a lack of organisational incentives and political will to use data. For instance, in countries where political and academic freedom is limited there is less space for using data for policy making purposes as policy outcomes are generally not challenged. Furthermore, many low income countries face major problems with accountability, participation, corruption and lack of political incentives to draw in evidence in policy implementation.

Likewise, there are factors that enable the effective use of data. Research shows that good data governance and infrastructure, organizational strategies that foster data use and the presence of feedback loops, lowering cost of data acquisition and high quality interactions among data users and producers--with clear incentives for both--all enable sustained data use. Partnerships and coalitions of stakeholders along the data value chain can also enhance the resilience of data practices and embed data use in everyday activities and strategies. 

The questions we’re asking on how to harness data for development

These themes of data use, barriers and enablers are our starting point for exploring pathways leading to sustained data use for development. Here at the Data Values Project our third track of action is focusing on data that delivers sustainable and equitable development.

Building on the diverse knowledge and experiences of the Global Partnership for Sustainable Development Data’s Technical Advisory Group, we are exploring pathways to sustained usage of data for development and identifying ways in which our collective advocacy can drive change. Particularly, we are looking to leverage collective knowledge and experience on the incentives and demand structures that drive sustained data use. We will explore models for scaling and institutionalizing these practices in public and private sectors and in civil society.

The initial phases of this track of action are focusing on four key questions:

1. What incentives contribute to sustained data use by different stakeholders, and how should such incentives be structured?
2. How can we foster increased and sustained demand for data from prospective users?
3. How do choices related to infrastructure, tools and methods affect potential for sustained data use?
4. How can we  harness local solutions that address local data users’ needs, specifically by examining measures to mitigate digital and data colonialism?

The Data Values Project will gather input from the data for development community on these questions to build a knowledge base of examples, evidence and best practices concerning sustained data use for development. We invite you to join the discussion,  participate in the track’s events and share examples, research or analysis that responds to the four questions we plan to tackle via our email [email protected]. 

Deepa Karthykeyan is the Founder and Managing Director of Athena Infonomics. Grant Cameron is Director of the Sustainable Development Solutions Network - TReNDS. Deepa and Grant co-lead the Data Values Project track on 'Data that powers sustainable and equitable development.

GPSDD’s Jenna Slotin and Martina Barbero provided input into this blog.

Interest in data governance is on the rise. The number of people worldwide searching for ‘data governance’ on Google has grown steadily since 2016. Data governance has also taken centre stage in the development space--the World Bank’s landmark 2021 World Development Report: Data for Better lives put it at the center of its analysis. 

Despite this growing interest, no operational or commonly agreed-upon definition of data governance exists. Perspectives on what it means for data to be well governed vary. The term “data governance” doesn’t appear in mainstream English dictionaries. Formal agency definitions tend to reference broad concepts like accountability, decision-making rights and power relations as well as practical concerns of data “processes and rules.” Google searches relating to data governance indicate that the term is associated with data protection and data management and that searchers often seek information related to the “benefits of data that is well governed.”  

The absence of a common understanding of data governance can lead to growing power imbalances over concentration of data benefits, lack of oversight from citizens and the public sector on data use, and can also result in perpetuating harms. Even our use of Google Analytics in this post highlights the questions of power that lie at the heart of data governance by relying on a private company that curates information without public oversight to measure public interest in data governance itself. These dynamics are rarely addressed by definitions that focus on practices, processes and rules. 

Conflicting perspectives on data governance in development 

This definition vacuum is starkly apparent in the data for development community, with practitioners and experts holding different views on the scope of data governance and what this means in practice. 

Current interpretations generally break down along three dimensions relating to conceptual definitions, organizational goals and the level of implementation:

1. Data governance includes technical and normative definitions. For some, it represents a technical concept related to data management, access rights and technical infrastructure for data management. For others, it encompasses normative elements such as the balance of power among actors and the distribution of accountability, rights and value related to data. For many, data governance includes a bit of both. 
2. The purpose of data governance varies by sector and community. Professional and sectoral communities define data governance through the lens of their unique goals, interests and mandates. For instance, international organizations tend to interpret data governance as including policies on personal privacy and cross-border data flows. Private companies on the other hand see it as a matter of managing access and creating accountability for data-driven decisions. 
3. Data governance manifests at different levels. Personal data governance applies to how individuals decide to share and use their own data.  Data governance also concerns how local and national governments manage complex data ecosystems. It also has a supranational dimension related to international trade, multilateral rules and agreements and-increasingly-to geo-political divides.

These differences emerged in public discourse precisely because of the success of the data revolution. We have seen remarkable progress and value in the use of new data sources for development. But as such innovations have shined a spotlight on the topic of data governance, many of the promises of the data revolution have not yet materialized. The development community is just starting to reckon with the negative externalities and direct effects of the increasing availability and use of data. In this context, unpacking what constitutes good data governance and how it can help deliver on the Sustainable Development Goals becomes not only necessary but urgent. 

Our approach to pushing reflections further

The Data Values Project is well-placed to navigate these diverse and often conflicting perspectives. The Global Partnership for Sustainable Development Data’s Technical Advisory Group members—particularly those leading a second track of action on Data Governance—are engaged on the front lines of activity and debates in data governance in international development today. 

Some of us come from the official statistics community and the UN Statistical Commission’s work on data stewardship. Our members also include activists building a movement to protect digital rights in the wake of COVID-19 and researchers studying governance frameworks and data responsibility. We represent civil society organizations experimenting with new models for multi-stakeholder data governance and companies establishing practical tools for ethical oversight and sustainable models for data sharing. 

Drawing on these and other connections, our track will bring diverse stakeholders together to shed light on what data governance means within different sectors, regions and communities. Our engagement so far indicates a need to explore others’ perspectives to identify areas of consensus and disagreement where more discussion, research and experimentation are required.

Through the Data Values Project, we will:

• Reflect on how good data governance can contribute to addressing, instead of reproducing, structural inequalities and
• Explore opportunities and priorities for change in current data governance approaches at the local, national and international levels.  

Building on areas of consensus, we are working to identify the specific elements necessary for good data governance. Ultimately, we will be advocating the changes needed to establish a clear mandate for better data governance. Do not hesitate to share your perspective on what data governance means and send your thoughts and contributions to [email protected]. 

Francesca Perucci of the United Nations Statistics Division, Frédéric Pivetta of Dalberg Data Insights, Camilo Andres Mendez Coronado and Karen Lizeth Chavez Quintero of the Departamento Administrativo Nacional de Estadística (DANE) and Tom Orrell of DataReady co-lead the Data Values Project track on 'Data that is well-governed'.

GPSDD’s Jenna Slotin and Martina Barbero provided input into this blog.
 

It’s time we asked ourselves some hard questions.

There is an extreme power imbalance within the development sector, derived from deep-rooted colonial attitudes and patterns of patronage. This inequality of power has resulted in an overriding norm that providers are good and are doing the job for the right reasons, while recipients should be grateful for any assistance they receive. This narrative must change, from one of providers and recipients to one of duty bearers and rights holders. As the wider development sector looks to tackle these entrenched power inequalities, the data for development community must reflect on some fundamental questions on the politics around data.

The 2030 Agenda for Sustainable Development is clear about the role of data in leaving no one behind. Ensuring that no one is invisible and that their experiences and needs are captured is essential for designing inclusive policies and programs. Without this commitment, history tells us that only some  of our communities and societies will progress. Often, those in the most disadvantaged situations will fall further behind. 

The focus on data is even sharper now than it was when Agenda 2030 was adopted in 2015 - as we’ve seen with the publication of the landmark World Bank Development Report on data this year. As we continue to grapple with the inequalities exposed and exacerbated by  the COVID-19 pandemic, the role of data has never been more important—to inform response, recovery and ultimately hold our decision makers to account.

But at the same time, data is no longer seen as universally beneficial, with governments and citizens around the world all grappling with challenges around privacy, misuse of data and the rising power of the Tech Giants. The application of data and digital technology in the development sector risks reinforcing  power imbalances and if left unchecked, data colonialism will threaten the efforts of the data revolution. Much of the data collected across the world is extracted from the communities and nations that it is about but  stored, used and disseminated to actors globally. Rather than data collected and used locally to inform direct service provision, data is frequently shared with those who have more power but are less accountable to the communities that data is about.  

Therefore, we are not just talking about more inclusive approaches that fill data gaps, we need to ensure full inclusion across the data value chain - from design and production to dissemination and use. The human-rights based approach to data is increasingly used across various settings; from official statistics to data generated by civil society organizations and communities themselves. Development planning to poverty measurement are two concrete examples, and demonstrate that being part of, and ideally having a leadership role across, the data value chain can empower people and communities to demand what they deserve and hold their leaders to account.

Ultimately it will be our values, our integrity and our collective ability to bring wide-ranging voices to the table to mitigate against power inequalities within our work that will be the test of whether data truly becomes a route to equality and inclusion.

Development Initiatives is proud to be co-leading track one of the Data Values Project on ‘Data as a Route to Inclusion and Equity,’ working alongside a diverse set of people and organizations from the Global Partnership for Sustainable Development Data’s Technical Advisory Group, along with fellow co-leads Al Kags of the Open Institute and Florentin Albu of the Children’s Investment Fund Foundation. This track is exploring the following questions:

1. Addressing inequality: In what ways can data and technology deepen or lessen inequalities, and what can be done to rebalance towards the latter? 
2. Translating data into practice: What are the factors that influence whether more inclusive data leads to more inclusive policies and programs?
3. Promoting inclusion: What does genuine inclusion look like across the data value chain and how do we make this standard practice?

Our journey so far

One of our first discussions was around language, including the language we are using as part of this project. Originally the title of our track was ‘Data that delivers for the most vulnerable.' This seemingly innocent language is loaded. It validates top-down power structures and implies that vulnerability is an inherent state, rather than a result of external actions. It reinforces the power inequality by setting the ‘data’ providers as givers while ‘the most vulnerable’ are recipients robbed of agency and context. If the Data Values Project is going to bring voices together to address the power inequalities around data, then we must first ensure that the language we use represents our values.

So we decided to rename the track as ‘data as a route to inclusion and equity.' This language emphasizes our core approach, to:

• Listen and be led by the voices of the communities we aim to serve;
• Act with groups and communities, rather than for them;
• And recognize the ability of data to support the agency and voice of communities. 

Throughout 2021, we will be listening to a range of voices as part of an open dialogue phase, hearing perspectives on what genuine inclusion looks like across the data value chain and how we can make this standard practice. We want to hear from you. We welcome your ideas, your challenges, and any resources you think might be helpful. Please get in touch via [email protected].

Claudia Wells is Director of Data Use at Development Initiatives and co-leads the Data Values Project track on 'Data as a route to inclusion and equity' alongside Florentin Albu of the Children’s Investment Fund Foundation and Al Kags of the Open Institute. 

The Global Partnership's Jenna Slotin and Karen Bett provided input into this blog post.