When a large percentage of the world population is under quarantine and with increased dependence on digital technologies –for work, exchanging with friends and family, making purchases for home delivery or organizing meetings online–, the issue arises once again of the prevailing model of development of these technologies. For the corporate giants that dominate the area, it is a golden opportunity to continue venturing deeper and deeper into our lives to extract the data that yields them profits.

Up to now, the concern that most data extraction occurs without the knowledge or authorization of those who provide it (and today that is just about everyone, bar those communities that are still isolated from technology) has meant that public debate mostly focuses on individual rights, privacy and security, in reaction to the regular reports of cases of abuse. Therefore, the solutions put forward tend to assume that such violations are exceptions and excesses which need to be regulated (some even argue for self-regulation by the companies involved).

This narrow focus, however, fails to address the broader picture, which is that of a digital economy where wealth and power is built precisely on the unlimited accumulation, processing and monetizing of all kinds of data generated by individuals and communities, extending far beyond the specific data that the company or entity requires for improvement of their service.  While a reasonable amount of data collection and processing can be an indispensable element for offering quality digital services (eg traffic control or health diagnosis), the fact is that today, almost every aspect of our lives, even down to our intimate thoughts, is being registered, interpreted, commoditized and marketed by these companies.  In other words, the violation of intimacy and the abuse of data is not an exception or an occasional excess, it is the very essence of the new digital economy.  Public scandals only break out when a security breach leads to the unintended revelation of a small part of this ongoing, but largely invisible, daily practice.

This situation is unprecedented in human history in its scope, speed of evolution and implications for power concentration, while people feel largely helpless to prevent it. Even the governments and legislators of the world’s major economies face severe limitations to do so. Given the unquestionable usefulness and allure of a large part of the technology, we tend to accept that it comes with the condition of ceding our data and we forget (or are unaware) that the present corporate model of technological development is not the only one possible. So the question we need to pose is: what would be the conditions that could enable a more people-oriented and less manipulative technological model to flourish?

It is this question that the Just Net Coalition (JNC) is aiming to answer with its recent Digital Justice Manifesto, launched at the Internet Governance Forum in Berlin, last November.  It proposes community ownership of data (other than strictly personal data) as a central principle for legislation and regulation, which need to be further broken down into specific rights and principles[1].  In order to have a better understanding of the rationale behind this proposal, as well as some of its more polemical aspects, ALAI held a dialogue, some weeks back, with one of the Manifesto’s authors, Parminder Jeet Singh, executive director of IT for Change (India) and founding member of the Just Net Coalition.

Addressing default corporate data ownership

Concerning why Just Net puts both personal and community data ownership at the center of the debate of the data economy, Parminder identifies two issues.  The first is that, while there is a lot of talk about security, privacy and personal data protection–which of course are very important–the bigger issue, that is rarely addressed, is the economics of data.  He explains: “We know that the most important asset is data - and even the mainstream press talks about data as the new oil, the new gold. In the feudal era, all economic relationships and all social and even cultural relationships were organized around the economics of land; and in industrialization these shifted first to ownership around industrial capital and later to intellectual property capital; and that determines to a great extent the distribution of power, of wealth, of opportunity.  In the digital age, this would happen around the economic resource of data, but sufficient attention is not yet paid to data as an economic resource. It is often felt that if we talk about data as an economic resource, we are kind of talking the language of corporations. But we can’t leave this as an open field. And when I say ‘we’, in this case, I’m talking about progressive civil society actors who work for the interests of the people at large. So just to say that we won’t talk about the economics of data is to leave the playground entirely to the corporations to make their own rules and establish default ownership around data.”

JNC therefore also focusses on the economics of data and economic and social rights around data, beyond the political and civil rights subsumed under a privacy (and freedom of expression) framing. So the reference to ownership has to do with who has primary economic rights over a set of data; and for that reason, JNC uses ownership as perhaps an imperfect analogy from the industrial age and feudal age to apply to data, while also recognising that data as an asset could involve different actors simultaneously having differential rights.

The second issue identified by Parminder is that–as many scholars have argued–ownership regimes do not necessarily come into existence only through legislation.  There are also default ownership regimes, such as occurred during the gold rush in the Americas, when gold mining was not allocated ownership, but people established default ownership by a balance of power based–in that case–on the respective gun-power of the different gangs.  “So now, although there is no law on ownership of data and people don’t want to talk about it, the fact is that the large data corporations consider the data that they collect as theirs; and then guard it behind technical walls as their de facto property.”  In conversations with company representatives about data, Parminder says they almost invariably come to a point where they claim that “this was data which we collected, we put effort into collection and therefore it’s our property”. Therefore “it’s evident that a default and highly concentrated data ownership has been established with respect to most of the society’s data, and such ownership is concentrated in very few hands–largely US companies and now some in China.” And the only way to counter this default ownership of data is to assign its legal ownership to the data source or the data subject, which is the people: individuals and the community.  Moreover, he warns that if we let this default ownership practice go on for much longer, it will become quasi law, and may even be accepted into law, because the digital relationships and structures would all be quite entrenched, and almost impossible to reverse.

Can data be ‘owned’?

Some civil society actors question the idea of ‘ownership’ or ‘property’ in relation to data.  A first concern is whether the nature of data make it possible to be ‘owned’; and secondly, whether establishing property rights over data wouldn’t play into the hands of the corporations that want to stake out their own rights over the data they collect, as they have done in the past with so-called ‘intellectual property rights’  So one question is whether there might be some other alternative to ‘ownership’, such as the ‘right to use’ of data…

Parminder Jeet Singh admits that, in the full sense of ‘ownership’, only physical objects can be ‘owned’, those things that ‘are yours alone to use’. He recognizes that, “with regard to intellectual property, most progressive civil society has been against this idea of owning ideas, when in fact the ideas those people may ascribe to a particular creator are generally coming from a much larger set of sources, from society.  But we have reached a situation where we have creators and creators have intellectual property.  But in the case of ‘ideas’, they are out there for anyone to share and copy, whereby corporations wanted law and ownership regimes to stop such free sharing. In case of data, it is already with corporations, behind its technology walls, and inaccessible to others, including those who contributed the data and to whom the data refers. Corporations do not need law in this case for exclusive appropriation. It is the data source and data subject people and communities who want a ‘data ownership’ regime to employ law to get access to their own data. With time, it is quite likely that another term than ownership might evolve that is found to be more appropriate for data, but for now ownership is a useful analogy.”

As for the term ‘right to use’, in fact ownership implies a mix of ‘right to use’, ‘right to license uses’, ‘right to control’, ‘right to regulate’, etc. Rights language is useful in that way, says Parminder; “but generally it also gets used to diffuse the issue, and deviate from data’s economic aspects, and their highly contested nature. And we currently want to sharpen the issue, highlight and focus on what we are trying to say which is primary economic rights to data.  The word ownership leaves no scope for anybody to think other than as primarily economic rights.  So we find this ownership word hard-hitting and it makes that point.”

From the content of the manifesto, it is clear that JNC is not talking about private rights.  “Here it is more in the sense of common property rights, as established by Elinor Ostrom and others: the word property implies ownership, even if as commons management,” Parminder clarifies, adding that eventual alternatives to the de facto situation should be fit to the purpose, whereas ‘right to data’ leaves things too diffuse. Even corporations are willing to share certain data, such as for public emergencies; but should the right to use be based only on their goodwill?

A further and justified concern is that property is normally considered to be alienable: something that one can wholly sell off; while many kinds of data rights, personal or collective, are not alienable: for example, a person cannot sell his or her privacy as such to somebody.  “This should be as illegal as it is to sell a person–or even oneself–into slavery. In the same way, there are certain kinds of data one should not be allowed to sell”: not only personal data, but also some kinds of community data.  Parminder recognizes this could potentially be a limitation of using the term ‘ownership’.  “We think that economic rights to many kinds of community data may be non-alienable, in the sense that a community may licence such data for certain uses or may even not allow others to use it, but many of those rights are unethical to completely sell off...” We have a very important political economy fight on our hands about data and we find ‘right to use’ is not an adequate term for that.”

Governing the data commons

So, what might a data commons and community ownership look like and how would it be governed? How would decisions be made?  India is the first country to employ the concept of ‘community data’,[2] and IT for Change, the organisation Parminder works with, has produced a paper proposing a framework law for data commons.[3] Parminder recognizes it will be a very complex exercise, “like asking at the start of the feudal age how land rights would be.” Land rights took decades and centuries to evolve, with various shades of them emerging, and this is still an ongoing process today.  “We are just presenting the first point of departure for similar, no doubt extended, explorations of rights to the key resource of the digital age which is data – which is the default ownership of most social data being with the community.”

Governance of community data being a complex issue, Parminder considers it should start with a framework law.  “There have been precedents: there is a Framework Water Directive of the EU which is about uses of water as a commons resource.  We need a framework law around the commons of data, which at a federal level lays out the enabling provisions for allocating economic rights and privileges related to data.  A key principle of such a framework will be: whether it is group anonymized data about the people, or it is data about natural phenomena associated with that community or artefacts like infrastructure which is associated with that community, all such data is community data and is by default owned by that community.”

The next step would be to define principles, such as how to determine the representatives of a particular community who act as trustees for the relevant community data and what powers they would have.  “In many cases it will be the local governments, or the interest-based communities may have elected representatives.” For example, Uber drivers in a city may form a council and elect union leaders to represent them.  These unions would have the power of making decisions over the data drivers collect while driving, which at present is unilaterally expropriated by the Uber company.  IT for Change has also proposed creating a National Data Trust, as a quasi-judicial body, which can inter alia adjudicate on appropriate community representatives or trustees for a given set of community data.  This is something that would need to evolve over a number of years.

Corporate use of data

A default community ownership regime for data would obviously impact how companies use data; so what might the implications be?  Parminder recognizes that if a company’s data assets are taken away, they may lose the incentive to keep collecting data that could serve the community.  There would be a need to establish kinds of rights and privileges to data, depending on different levels of processing undertaken. There is a difference between raw data, on the one hand, and data that has been processed to different levels and the inferences derived from it, on the other.  Parminder considers that, in the case, for example, of real-time city commuting data collected by a company, the law should enable the city to ask for that data to be shared, because it remains a community asset in its first-level formatting.  Then with the different kinds of processing, there would need to be regulations that clarify till which level is the data still a community asset that has to be shared and at which point it becomes a private asset; also, which kind of data may be part community and part corporate and which kind of data has to be put in open regulated markets.  “So there are many ways and depending on what kind of processing has been done by the corporations, they may have certain data related privileges so as to incentivize them to keep collecting it.”

Also, some of the data sharing obligations could be by uses. “There are two kinds of larger use cases: one is sharing for directly public interest purposes.  Every area of public policy making would soon be dependent on platform data: agriculture, health, education, you name it.”  But another kind of use relates to ensuring that local domestic industry gets access to data it needs to establish and develop their digital businesses, without which they will never be able to compete with the transnational corporations.  Countries such as the UK, France, India are already envisaging policies along those lines.

There would be differential treatment of data according to the sector involved; for example, the public interest concerns relating to health data would need different regulations than, say, general e-commerce data.

While India is the first developing country working on policies in this area, there is growing interest in other developing countries, such as South Africa.  In multilateral body discussions, the issue of data ownership frequently comes up.

In summary, JNC’s concept of data ownership arises from two considerations: people’s rights versus the power of corporations, and the geopolitical concerns of developing countries’ position in the emerging global digital value chains.  Currently, the default is that data ownership is concentrating within a few corporations, largely in the US, and now some in China. To respond to these concerns, says Parminder “we found that there is no other way than to start an opposite process whereby the community of origin of data claims the data: that addresses both issues of people’s claims and rights vis a vis corporations, and also the geopolitical issue.  These are the reasons which took us to the ‘data ownership framing’, because when you have real problems you need responses.”

In these times of the pandemic, proposals such as these take on renewed relevance, since there can be no doubt that the digital corporations are taking advantage of the present situation to further strengthen their model of extraction and appropriation of our data.

- Sally Burch is a British-Ecuadorian journalist, executive director of ALAI and a member of the JNC steering committee.

@SallyBurchEc