Data Cooperatives for Pandemic Times

Image Credit: Artwork by James Seibold, Public Domain.

When COVID-19 hit populations worldwide, it became clear how deeply data practices are connected to democracy. One recent report, for instance, found a ‘data divide’ highlighting inequalities in access, knowledge, and awareness of digital health technologies used in the pandemic. The long-researched ‘digital divide’ is also a ‘data divide.’

In response to unjust data practices, we propose that the economic resource of data should be organized locally through data cooperatives.

Data cooperatives are member-owned data management systems. They can limit or modulate the capture of data to create value for their members in areas of application where the affected communities see no apparent benefit. The data can only be used following the rules that members approved. Data co-ops hold fiduciary obligations to their stakeholders. Local data storage may encourage communities to make collective decisions about the collection and use of their data. Communities can use the revenue from their data to bring about positive social change by funding public research, for example. Data co-ops are part of the ecosystem of the digital cooperative economy. 

Among scholars, co-op founders, civil liberties defenders, legal scholars, mayors, and technologists, discussions about various forms of community ownership of data have intensified. The lexicon of this debate includes keywords like data portability, data stewardship, data sovereignty, the data commons, and cooperative digital infrastructure.

The leadership of the city of Barcelona, for example, sank in its heels, demanding technological sovereignty over the data that companies like Uber collect when operating in the city. A group of British scholars advances the work on data trusts. In New Zealand, data practices privileged the dominant NZ European Pākehā population, leaving indigenous Māori communities with unequal access to data about COVID-19. In Canada, equally, unjust data practices led First Nations to demand indigenous data sovereignty. 

Ownership Matters

Who would be ideologically better suited to collectivize data than cooperatives? In principle, cooperatives should be terrific at sharing data. After all, one of the principles of the International Cooperative Alliance (ICA) is ‘cooperation among cooperatives’:

Cooperatives serve their members most effectively and strengthen the cooperative movement by working together through local, national, regional and international structures.

But also modern corporations understand the power of cooperation. The ‘co-opetition’ between Apple and Google is one example. In exchange for building Google’s search engine into its products, Apple receives $8-12 billion dollars a year. That amounts to 14 to 21 percent of Apple’s annual profits. Corporations also fund think tanks to provide an intellectual rationale for their activities. By any measure, currently—and paradoxically—corporations are far better at cooperation among themselves than the cooperative movement is. 

Beyond the cooperative movement, pioneers in the digital cooperative economy are also inspired by Catholic values, African Ubuntu philosophy, and the subsistence economics at the heart of Buddhism. All three make the common-sense argument for sharing and interconnection. 

But in countries like the U.S., exchanging data among co-ops may be perceived as the formation of cartels, prohibited through the Sherman Antitrust Act of 1890. Regulatory intervention should impose, as part of antitrust action, the mandate for all platforms to allow for data portability. This would soften the network effects that keep the five largest tech companies unchallenged.

Data cooperatives are part of the landscape of people experimenting with community ownership in the digital economy. But why does ownership of digital platforms or protocols matter at all? The ownership of assets has profound implications on many aspects of life; it impacts income inequality and, concretely, the daily experiences in the workplace. With cooperative ownership, members jointly decide the rules of the game. Responsibly collecting, storing, and sharing data is key to unlocking its potential. Governance of such data means that the data rights of communities are protected, and the use of data is fair and focused on social benefit. It changes the current paradigm of individuals surrendering data to large tech companies to a community ownership system. Communities retain data rights, accountability, legal standards, and fiduciary representation. With data co-ops, data collection focuses on what is beneficial for communities. They decide which data are collected about them and for what purpose. To avoid mission creep, data cooperatives need to adhere to the globally shared definition of what makes an organization a cooperative: 

A cooperative is defined as an autonomous association of persons united voluntarily to meet their common economic, social, and cultural needs and aspirations through a jointly-owned and democratically-controlled enterprise. 

This definition matters as it emphasizes that cooperatives are autonomous and organized voluntarily to meet local needs. Imposing the cooperative model is unhelpful and possibly harmful.

Currently, in the crypto and token economy space, W3 prototypes of data cooperatives are seeing the light of day. Blockchain and smart contracts allow those with access to these technologies to create tokens that can be used for the governance of co-ops. But so far, despite interesting ongoing experiments, how to design and apply token systems is still unclear. In the crypto space, we find unfortunate misunderstandings such as the claim that VISA and SWIFT networks are genuine cooperatives. 

The Data Co-Op Pioneers

To be sure, data co-ops are not a figment of the academic imagination; they are already among us. Representative cases include Salus, Driver’s Seat, LBRY, dOrg.tech, and MiData, Gooddata, Mnemotix, and Polypoly.

Salus Coop is a non-profit data cooperative for health research (referring to health data and lifestyle-related data more broadly, such as data that captures the number of steps a person takes in a day), founded in Barcelona in September 2017. Salus aims to create a citizen-driven collaborative governance model and management of health data. It legitimizes citizens’ rights to control their health records while facilitating data sharing to accelerate public research innovation in healthcare.

MIDATA is a ‘health data cooperative’ started in 2015. MIDATA.coop enables citizens to securely store, manage and control access to their personal data by helping them to establish and own national/regional not-for-profit MIDATA cooperatives. MIDATA cooperatives act as the fiduciaries for their members’ data. MIDATA offers a platform on which user-members can securely store copies of their medical records, genomes, and mHealth data. Members might decide to give their physicians access to all personal data through the platform. In contrast, a not-for-profit cancer research institute could be given access to only medical and dietary information. Members could deny access to a for-profit drug company. Members’ revenues from the sale of data are donated to public research.

LBRY states: “We think users should own their content (and their privacy) instead of handing it over to a corporate giant and their advertising buddies. If you think we’re paranoid, there are dozens of examples of companies abusing users and acting against their interests. It’s not paranoia if they’re actually out to get you.”

Polypoly is a data cooperative that ensures that personal data no longer leaves a device, whether mobile phone, computer, or web-enabled toaster. PolyPod, the member of the co-op, has a private server that stores his/her data, and that is controlled by him/her.

dOrg.tech, currently transitioning to a data co-op, is a full-stack development collective that works with industry-leading projects in Web3 by using a decentralized manner by builders worldwide through smart contracts, blockchain, and Ethereum.

Data cooperatives anchored in local communities shaped by their specific cultural traits and features could become a practical pathway to resisting data colonialism.

But of course, there are obstacles. Communities will have to be educated about the cooperative principles. Beyond that, such local data cooperatives may not be easily replicated because they are rooted in their contextual and territorial critical factors. Over time, also governance within such cooperatives may become a challenge: they could erode or be destroyed as some may people join and others may have to be asked to leave.

We are not mistaking data cooperatives for a one-stop-shop that will fix all the ills of platform capitalism. We understand data cooperatives as an approach that is part of a toolbox that also includes efforts such as Solid, but also unions, neighborhood associations, regulatory intervention, public ownership, and hybrids among these forms.

We also acknowledge the limits of the framing of data ownership. What does data ownership or ownership of a digital platform mean when most groups rely on commercial upstream services? What does it mean when data cooperatives are using proprietary software and the cloud services of large tech companies? An alliance of cooperative data services offered from the margins might gradually overcome commercial upstream services by decentralizing data governance models at the local level.

Unexpectedly, in Law and Business Schools in the United States and the United Kingdom—traditionally, the places where you’ll find resourceful experts of the modern corporation—there is much interest in data cooperatives right now. This is surprising given that the co-op model is a well-kept secret in the halls of most of these schools given that faculty neither teach nor research the cooperative form.

When working on data co-ops, legal scholars and technologists should be aware of the colonial history of cooperatives that left an acidic aftertaste for many people. British colonialists championed cooperatives in the fishing villages of Hong Kong and the Indigo fields all over India. Colonial powers supported the introduction of cooperatives to advance their economic interests. In Africa, colonial powers steered cooperatives primarily towards the production of cash crops for their home countries, namely coffee, cocoa, tea, and cotton. In colonial South Africa and Kenya, agricultural cooperatives owned by Whites were heavily subsidized by governments. This model incentivized European settlers while also making sure that co-ops had monopolistic control of these sectors.

Earlier, we had emphasized that co-ops are autonomous. While cooperatives of any kind cannot thrive without government regulation, they should only set enabling conditions. Governments in Kerala, India, and Sri Lanka govern cooperatives top-down, which directly conflicts with their core mission.

Co-op whitewashing can generate great harm. Brazil is a historical example of what can go wrong when lawyers take the lead designing the legal framing for cooperatives. In 1994, Art. 442 of Brazilian Labor Law mandated that “Whatever the branch of activity of the cooperative society, there is no employment relationship between itself and its members, …” As a consequence, labor rights were harmed as many businesses now incorporated as co-ops simply to avoid having to pay benefits to workers. 

‘Nothing About Us Without Us,’ the slogan of the international movement of people with disabilities, also fits into the context of the creation of legal and technical templates for data cooperatives. Communities, not academics or governments alone, should shape the legal framing of data trusts, data cooperatives, and cooperative digital infrastructure.

One of the technical architects of the General Data Protection Regulation (GDPR), Professor Alex Pentland at MIT, called for data cooperatives to become the guardians of community data. According to Pentland and Hardjono, with 100 million people members of credit unions in the US, the opportunity for community organizations to leverage community-owned data is massive. Research has shown that the various types of co-op models generate very different mutualistic benefits. What makes credit unions or consumer co-ops the go-to model for legal scholars and technologists? The multistakeholder model, based on the research, would seem a more likely candidate.The credit union model has not shown the strongest results when analyzed through a social justice lens. While being very large, credit unions have been criticized for straying from the cooperative principles, low participation, and engagement, frequently not offering a distinct, felt difference to conventional banks. We also question the objective to create dominant data cooperatives through credit unions for community data. Cooperatives, when acting as de facto monopolies, have not shown to behave meaningfully better than other organizational forms.    

At the very least, communities should be at the table when discussing legal instruments for data co-ops. Co-op scholars, co-op historians, and leaders of co-op organizations, especially from LGBTQI communities, indigenous peoples, and those from the Global South, should be part of the conversations about legal templates and technological infrastructure for these data institutions. An alliance of data services offered from the margins might help to diversify the digital economy gradually.

We welcome the embrace of the co-op model and understand the thrill of pioneering legal or technical models for novel data management. Proposals for data co-ops, however, should be co-developed with co-op practitioners and connected to the long history and analysis of the various forms of cooperatives. This includes the history of people working toward a cooperative commonwealth, based on the vision of a society rooted in cooperative and democratic socialist principles. A cooperative commonwealth is comprised of a network of small, autonomous, interlinked but competing co-ops. Just think of the Knights of Labor that fought for a multiracial ‘cooperative commonwealth’ as early as the 1880s.

To serve such a vision of a digital commonwealth, localized data need to be federated into data ecosystems. Distributed ledgers are likely able to help. Communities will be using their data gathered in local repositories. This is the case with Eva.coop, a Montreal-based data cooperative: It provides the digital infrastructure for cooperatives of drivers without accessing local passenger data.

Eva.coop is built on the EOSIO blockchain protocol as a way to show how the cooperative model could mark a new blockchain-based iteration of the ‘sharing economy’ driven by a democratically centralized system that respects the privacy of workers and meets local needs. Local communities have more input, and drivers are treated more fairly, riding members maintain their privacy, and comforted by a locally supported app. Such federated, democratically centralized data ecosystems can be arranged by sectors (e.g., health-related data, environmental data, transport, and mobility data, energy and consumption data). Communities can then decide to release those data meaningfully. 

Without such bottom-up collaboration and clarity about the cooperative principles, there is a considerable danger of ‘co-op whitewashing,’ the propagation of fake platform co-ops and the abuse of the cooperative model. Such ‘co-op whitewashing’ would be ironic given the fact that over the past decade, the ‘sharing economy‘ employed a deceptive discourse that used the language of counterculture, love, and cooperatives to market commercial platforms.   

The COVID-19 pandemia will hardly fade away in the upcoming months or even years. More urgently, it has already revealed great vulnerabilities. Reshuffling the ‘old normal’ will not be enough. An internationalist movement inspired by a vision of a digital cooperative commonwealth is required to structure a social and economic post-pandemic transformation. We consider these pandemic times to be a period of transformation, an opportunity for experimentation with data co-ops, working toward more responsive systems of care, adequate medical services, and a more participatory, fair data economy. Data cooperatives must be shaped by those who need them most, grounded in the history and practices of the global co-op movement.