- A major challenge in thinking about a Canadian “data strategy” stems from the fundamental tensions between what information “wants” to be: free and shared, but also expensive, owned and controlled. Sometimes, information wants to be dangerous.
- The law of the sea presents a powerful and useful analogy for thinking about the international governance of data.
- Reflecting similar tension between what the sea “wants” to be, the law of the sea reflects several important functions that the sea has performed: as a medium of communication; as a reservoir of resources; and as engaging matters essential to the state’s security, or that otherwise involve political and strategic considerations.
- Developed in the course of geopolitical struggles over the sea and competing claims for ownership and control against claims for openness and freedom, freedom of the seas prevailed and established commons governance of the sea as the default principle, as well as the conditions for deviating from it.
hirty years ago, Stewart Brand (1989, 202) famously observed two simultaneously conflicting but accurate truths about information: “Information wants to be free. Information also wants to be expensive.”
Information wants to be free, he explained “because it has become so cheap to distribute, copy, and recombine — too cheap to meter” (ibid.). Information wants to be expensive (and owned) “because it can be immeasurably valuable to the recipient.” This tension, which fuels “endless wrenching debates” about the governance of information and practices surrounding it, will not go away, “because each round of new devices makes the tension worse, not better” (ibid.).
A major challenge in thinking about a Canadian “data strategy” stems from this fundamental tension. And if that is not challenging enough, Joshua Gans (2012, 29) reminds us that what information really wants is to be shared. Information wants to be shared “because it is often the case that when more people use or consume some piece of information, an individual’s value of its use and consumption rises.”
To complicate things further, not all information becomes more socially valuable when shared. The consumption value of baseless rumours, fake news and other falsehoods might increase with sharing, but could inflict various types of externalities on society. Moreover, sometimes information “wants" to be dangerous. Information about individuals’ vulnerabilities could be exploited to harm them, and it is better if some information concerning national security is not shared. Information can also be politically dangerous: it can inform and empower the powerless and the marginalized and help challenge and disrupt existing power structures, or it can be used by the powerful and the privileged to surveille, control, manipulate, oppress and dispossess the poor, the weak and the marginalized. Information of this type might “want" to be regulated: sometimes for good purposes, sometimes for nefarious ones, depending on the regime.
Our existing laws treat various types of information differently: some information is free, other information is expensive; some information is shared, other information is owned; the dissemination of some information is unrestricted and encouraged, while in other cases it is discouraged and supressed; some information is public, while other information is treated as private, privileged or secret. Nevertheless, restrictions on access to and dissemination of information are the exception and freedom is the norm. Or at least, this is what we expect in a constitutional democracy. Indeed, the freedom to disseminate information and the right to receive it are constitutionally protected under the Charter of Rights and Freedoms,1 and therefore could only be restricted “by law as can be demonstrably justified in a free and democratic society.”2 Moreover, even when the law imposes restrictions on the dissemination of information, such as in the case of copyright in expressive works, “there can be no copyright in ideas or information, and it is no infringement of copyright to adopt the ideas of another or to publish information derived from another.”3
The freedom to disseminate information and the right to receive it are constitutionally protected under the Charter of Rights and Freedoms.
Yet, data is said to be the essential capital stock of the data-driven economy, built around massive data collection and various business models for profitably sharing and using it. Metaphors such as “the new oil” or “the new gold” reflect this value extraction potential for businesses and they conjure up the “wants to be expensive” theme. These metaphors emphasize the money that can be made by those who control data — the private benefits that they might derive from its exploitation, not the aggregate value shared by society as a whole. Such metaphors imply ownership and exclusive control (we do not hear as often that data is “the new air,” “the new light” or “the new water” — resources much more valuable than oil or gold, but which, for the most part, are governed as commons, and “want to be free”).
But “metaphors in law are to be narrowly watched, for starting as devices to liberate thought, they end often by enslaving it.”4 Choose the wrong metaphor to drive your strategy, and you get failure or even disaster.
Thomas Jefferson, in one of the most famous documents in the history of intellectual property, wrote that an idea (and this would equally apply to data) cannot be susceptible of exclusive control, because an idea has the “peculiar character” that “the moment it is divulged, it forces itself into the possession of every one, and the reciever5 cannot dispossess himself of it.…no one possesses the less [of an idea], because every other possesses the whole of it” (Looney 2009, 383). Jefferson had served as a member of the US Patent Board and was quite aware that a patent can be very valuable to its owner, but he insisted that ownership and property were not the right way of thinking about these policy issues. Instead, he preferred another metaphor: “He who recieves an idea from me, recieves instruction himself, without lessening mine; as he who lights his taper at mine, recieves light without darkening me” (ibid.).
So, consider another metaphor: data as “the new sea.” Unlike oil or gold, but like the sea, data and information are non-rivalrous resources that can be used simultaneously by everyone without being diminished. And the law of the sea is built around similar tensions between what the sea “wants to be”: free, shared and open; expensive and owned; dangerous and controlled.
“Freedom of the seas” is a cornerstone principle of international law, but this has not always been the case. During the fifteenth and sixteenth centuries, Spain and Portugal proclaimed the “closed seas” concept, supported by the Papal Bulls of 1493 and 1506 dividing the seas of the world between the two powers (Shaw 2008, 609). Spain and Portugal asserted that because they discovered new navigation routes to territories in Asia and America, they also “owned” the right to trade with those territories and were entitled to exclude other nations from trading in and with those territories.6
The Dutch, a middle power with big trade aspirations, challenged those claims. After the seizure of the Santa Catarina, a Portuguese merchant ship, by the Dutch East India Company, the company asked the Dutch jurist Hugo Grotius to develop a counter-argument in favour of the freedom of the seas. Grotius’s Mare Liberum established the principle that international waters are treated as commons, “accessible to all nations but incapable of appropriation” (ibid., 554).
In denying the Portuguese claims, Grotius disputed the view that the high seas could be owned. He insisted that the sea “wants to be free” because it can be used by one person without lessening the use of another. Accordingly, the sea, like “all things which can be used without loss to any one else” ought to remain in perpetuity for the common use of all people (Grotius 1916, 28).
Like Jefferson’s discussion of exclusive rights in ideas two centuries later, Grotuis compared a person claiming a right to exclude others from navigating the seas to the person who “should prevent any other person from taking fire from his fire or a light from his torch” (ibid., 38). Such a person should be accused “of violating the law of human society, because that is the essence of its very nature, as Ennius has said: 'No less shines his, when he his friend’s hath lit'” (ibid.).
The freedom of the high seas became a basic principle of international law, yet like most basic legal principles, it is not absolute, and a coastal state could still treat a maritime belt adjacent to its coastline, known as territorial waters, or territorial sea, as an (almost) indivisible part of its domain (Shaw 2008, 554).
Mare Liberum and the law of the sea present a powerful and useful analogy for thinking about the international governance of data. Grotius’ freedom of the seas principle prevailed over a powerful competing narrative seeking to justify exclusive rights for trading over the high seas, and established commons governance of the sea as the default principle, deviation from which requires justification. Likewise, the line of argument that Jefferson articulated established commons as the default governance structure for information and prevailed over a powerful narrative seeking to establish ownership of information as the baseline norm. In both instances, what should remain as commons and when deviating from this baseline might be necessary or justified remains a live question as new technologies and business models for extracting value from such common resources emerge, and as society’s needs and the problems it faces evolve.
The problems that the law of the sea encountered and the solutions it provided offer a useful framework for thinking about solutions to similar problems that may arise in the governance of data.
The development of legal rules concerning the seas reflects several important functions that the seas have performed: first, as a medium of communication; second, as a reservoir of resources (ibid., 553); and third, as in the case of territorial waters, as essential to the state’s security, or otherwise involve political and strategic considerations. Through its development, the law of the sea confronted the question of whether the sea wants to be free, owned, shared or regulated. More precisely, when do we want it to be free, owned or shared, and how do we want to regulate it? Therefore, the international governance of the seas, which includes a mix of different modes of governance, provides interesting examples that might be helpful in thinking about the governance of data.
Unlike the internal waters of the state, which are fully within its unrestricted jurisdiction, the coastal state’s sovereignty over its territorial waters is subject to the right of others to innocent passage. Still, the state may exclude foreign nationals and vessels from fishing within its territorial sea and from coastal trading, and reserve these activities for its own citizens, and it has extensive powers of control over matters such as security and customs (ibid., 570).
The law of the sea in the state’s territorial waters reflects the three aspects of the sea: innocent passage preserves the sea’s function as a medium of communication and, for this purpose, the sea remains open and free as it generally “wants to be.” When passage is no longer “innocent,” i.e., where it is “prejudicial to the peace, good order or security of the coastal state,” the state may exercise its jurisdiction and prevent it. Examples include “prejudicial passage such as the threat or use of force; weapons practice; spying; propaganda; breach of customs, fiscal, immigration or sanitary regulations; willful and serious pollution; fishing; research or survey activities and interference with coastal communications or other facilities” (ibid., 571). This reflects the recognition that, in certain cases, the sea “wants” to be controlled. The power to exclude foreign nationals and vessels from fishing, research or survey within the territorial sea also reflects the sea’s function as a reservoir of resources, some of which “want to be expensive and owned.”
While the departure from the freedom of the high seas principle in territorial waters was originally linked to the coastal state’s ability to dominate its territorial sea by military means, coastal states may now exercise particular jurisdictional functions in the contiguous zone, and international law has moved to recognize even larger zones such as fishery zones, continental shelves and exclusive economic zones in which a coastal state may enjoy certain rights to the exclusion of other nations. At the same time, there has also been a move toward proclaiming a “common heritage of mankind” regime over the seabed of the high seas (ibid., 554-55).
This governance of the sea can be a useful model for thinking about the international governance of data, as it provides a rich set of governance models dealing with the different aspects of the sea: some aspects are governed as commons, others as a shared resource, while others are governed as semi-commons, or subject to exclusive jurisdictional control. These models can be instructive not only where there are similarities between data and the sea, but also where there are differences.
For example, while the sea as a medium of communication is generally non-rivalrous and therefore amenable to commons governance as far as passage is concerned, the resources that can be extracted from it often are not. Thus, oil and gas are rivalrous resources and cannot be governed as commons, while fishing might seem suitable for commons governance in the short-run, but the risk of overfishing and the resultant “tragedy of the commons” might justify other governance models. By contrast, as a resource, data is non-rivalrous and therefore a strong case exists for insisting on using commons as the default mode of governance, while the means of communicating data might not be suitable for commons governance.7
These models can be instructive not only where there are similarities between data and the sea, but also where there are differences.
The governance of data requires rules with respect to the means of communicating data as well as rules about data itself. One set of rules might be comparable to the right of innocent passage. The law of the sea recognizes that within the territorial waters, complete commons governance may be neither possible nor desirable, yet as far as innocent passage is concerned, the law stops short of exclusive control, and even when the sea ends, the law of many states has for centuries required that the ports, and its internal navigable waters, highways and, later, railways and the mail, are open to all on a fair, reasonable and non-discriminatory basis, even when they are privately owned and “want to be expensive.”
Domestic law, including constitutional norms regarding privacy and search and seizure, as well as telecommunication policies, including “net neutrality,” have adopted similar rules to the transmission of electronic data. However, the same does not apply when data crosses the border. For example, as Andrew Clement (2018) notes, more than 80 percent of Canadians’ internet traffic is estimated to pass through the United States, making it subject to unprecedented surveillance. Even worse, when data about Canadian persons crosses the Canada-US border, it falls into a constitutional black hole: the Canadian constitutional position is that the Canadian Charter of Rights and Freedoms does not apply to extraterritorial searches and seizures, and the US constitutional position is that the Fourth Amendment does not apply to nonresident aliens (Austin 2016, 472).
In articulating his opposition to the Portuguese monopoly of the commerce with the East Indies, Grotius maintained that commerce should be open to the people of every state. While focusing on the sea, his argument was based on a broader notion of freedom of trade, which itself was only a manifestation of a broader concept of a universal human society, according to which all human beings constitute a universal human society, governed by common rules of law (ius gentium) applicable to all human beings and guaranteeing them fundamental rights (Ito 1974, 2). One of those fundamental rights is ius communicationis, the right of all human beings to communicate freely with each other (ibid., 4). For Grotius, unimpeded and peaceful maritime navigation was merely an extension of that right (Borschberg 2011, 83).
Grotius described the right to communicate in words that still resonate today: “God,” he wrote, “had not separated human beings, as He had the rest of living things, into different species and various divisions, but had willed them to be of one race and to be known by one name;…He had given them the same origin, the same structural organism, the ability to look each other in the face, language too, and other means of communication, in order that they all might recognize their natural social bond and kinship” (Grotius 1916, 1-2).
Still, this humanistic message was written in the context of a conflict between European nations over the conquest and colonization of other nations and peoples. In defending the freedom of the seas and the freedom to trade with the East Indies, Grotius articulated doctrines that preserved the freedoms and privileges of the powerful and technologically advanced colonizers, not so much those of the colonized. As Jonathan Obar and Brenda McPhail (2018) explain, the rules that we write about the governance of data may also empower the powerful and punish the marginalized.
Thus, the new oil or new gold metaphors might nonetheless be useful by reminding us how the quest for the riches that may be found in other territories has often resulted in the brutal devastation of the human beings and the communities that inhabited them — and warning us against designing data governance rules leading to new versions of plunder. However, the history of the law of the sea also reminds us that even rules articulated in the most humanistic terms might result in serious inequities. Much depends on the identity and interests of those who set the rules, and the processes of designing them.
Finally, while rules concerning data governance have increasingly become integrated in international trade agreements, trade agreement negotiations might not be the proper venue for developing the right set of rules. In fact, so far at least, trade agreements have entrenched a wrong set of rules.
On the one hand, instead of committing to a commons baseline for the treatment of information and data and making it easier for states to resist the constant demands of special interests who wish to monopolize it, intellectual property rights have expanded relentlessly through international trade agreements, despite the lack of evidence supporting the claim that such expansion would contribute to greater innovation, productivity or growth (Katz 2017). As a result, even if information wants to be free, more and more of it is owned and locked down, and it is more expensive than ever before.
On the other hand, recent trade agreements, such as the Trans-Pacific Partnership — now renamed the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) — include restrictions on the state’s ability to mandate data localization and otherwise regulate data transfers (Geist 2018). The CPTPP also prohibits mandated open-source policies, even though, in many cases, access to the source code of software could be an effective way to detect and fix software flaws “that may once have been capable of crashing applications [but now] have the potential to crash cars, planes, medical devices, appliances, and other connected infrastructure” (Claburn 2017). This prohibition also makes it more difficult to counter explicit and implicit biases that are coded into the growing algorithmic decision making. While the CPTPP requires its parties to adopt or maintain consumer protection laws and “a legal framework that provides for the protection of the personal information of the users of electronic commerce,”8 it does not include any specifics about the content of such frameworks or define any mandatory minimum standards for such consumer protection and privacy laws.
The combined result of the current treatment of information and data governance in trade agreements exhibits the worst of all worlds: they have imposed ever-growing restrictions on individuals’ ability to use information — the common heritage of mankind — while increasingly restricting states’ ability to address those instances where information can be misused and harm their citizens.
This should not be that surprising since modern trade agreements are less about free trade than they are agreements to manage their members’ trade and investment relations on behalf of each country’s most powerful business lobbies (Stiglitz and Hersh 2015). This makes trade negotiations a very poor venue for designing rules for the good governance of data.
Information can be free, shared and open. It can be owned, closed and expensive. Information can be empowering and it can be dangerous. Ultimately, it is up to us, as a society, to determine what we want it to be.
I wish to thank Shamnad Basheer for teaching me about Hugo Grotius’s Mare Liberum. See Basheer (2017a; 2017b).