Digital constitutionalism: the governance of virtual communities, part 1

I am finally beginning to write up my thesis. What follows is the first half of the argument I plan to present. This will hopefully provide the structure for identifying the problem and the context of the argument.

Comments welcome – what have I missed so far?


The internet provides the medium for a wealth of virtual communities, each with its distinct set of norms and values. Individuals all around the world participate in these networks to play, to socialise, to learn and teach, to express themselves, to do business, to communicate with friends and loved ones, to engage in political discourse and political process, and for innumerable other activities. These communities provide not only a 'space' for people to interact but also the promise that each individual will be able to find a community whose norms and values align with her own. Since the internet became popular in the 1990s, this libertarian idealism has proved extremely powerful. It argues that freed from physical scarcity and spatial barriers, individuals will finally be able to choose to associate with other, like-minded, individuals, and these communities will be able to determine the rules which best fit their society, rather than the clumsy approximations that often result from national democracies.

The main problem with the governance of virtual communities is that our legal system operates in such a way as to vest overwhelming power in the hands of those who create and maintain the platforms. These people, whom we call 'proprietors', for they own the code that defines the platform and servers upon which the code runs, exercise almost complete discretion as to who may access and who may continue to access 'their' community. The law, by giving primacy to these property rights, marginalises the interests of participants in these communities. Further, by casting any disputes or tensions which arise as belonging wholly in the 'private' sphere, we deligitimise any change to the current allocation of entitlements.

The result so far has been that the technologically deterministic claims of the cyber-libertarians have not held up, at least not in the largest virtual communities. The suggestion that individuals will vote with their feet (or their wallets) and choose to leave communities that do not reflect their own values breaks down as individual communities become more important. Network effects act to restrict both entry of competitors and exit of participants. As the value of many communities is proportional to the number of participants, new communities have trouble reaching a critical, sustaining, mass. Participants are less likely to leave an established community for a fledgling community, even if that smaller community has more appropriate norms or values. This means that proprietors do not have to be very responsive to the demands of participants in order to retain their custom.

There are, no doubt, limits on the behaviour of proprietors. A proprietor who is not responsive enough to the demands of the community will, eventually, begin to lose participants. In the various commercial models, this usually means a drop in subscriber revenue or in revenue derived from advertising. The relationship between participants and proprietors is accordingly seen to be market-based, and proprietors have an incentive to be just responsive enough not to lose too many participants. This model, in practice, is far removed from the idealism of the cyber-libertarians. Governance is essentially reduced to business decisions about the most profitable way to manage the virtual community.

There is a significant problem when the interests of participants are reduced to a market rhetoric. The activities of participants in virtual communities cannot be understood as the activities of mere consumers of entertainment product, and treating them as such leads to substantial injustices. For participants, there is much more at stake than access to a service – that access underpins their ability to communicate with friends and family, to express themselves, to carry out trade and commerce, and to participate in political discourse. Thus, for example, when a participant is threatened with expulsion from the strongly heteronormative World of Warcraft for advertising for a guild that is friendly to those with alternate gender or sexual identities, she risks losing not only access to a recreational pastime, but access to her rich social networks, her personal identification with her avatar, her virtual possessions, and the rest of the benefits that attach to participation in the community.1)

Similarly, when an individual is banned from a social networking site, she loses a significant ability to connect with her friends, family, and distant associates. The more effective the social networking platform is at changing the way that groups organise events and remain in contact, the more acutely she will feel this disconnection. The same disconnect occurs when an individual is denied access to cloud computing platforms and is no longer able to access her email contacts or stored documents.

In the corporeal world, these concerns are often seen as public concerns, and constitutional and administrative law principles have developed to restrain states from arbitrarily or capriciously taking away the ability of individuals to be secure in their property, their ability to communicate, and their freedom of association. In the privatised environment of the internet, however, there are very few guarantees. The limits of a proprietor's power are established by the bounds beyond which certain actions will become unprofitable. At its worst, this model approaches the worst failures of majoritarian or populist rule, where individuals and minority groups are often subject to harsh treatment and discrimination.

It follows that as the internet becomes more vital to the ways in which we communicate, do business, express ourselves, live, love, and learn, then the risk posed to individuals and groups dramatically increases. The law does not currently have the adequate vocabulary to deal with these risks. Constitutional action, particularly in Australia, provides limits on governmental action, but provides no individually assertable rights against private actors. This negative model assumes that, in the absence of governmental action, citizens will be free. To the extent that this model was ever accurate, it is certainly flawed in a context where access to crucial social networks and forums for self-expression requires access to another's 'property'.

In this context, the distinction between positive and negative restrictions becomes confusing. In the absence of particular government interference, we are left with the base rules of property and contract, which are nonetheless public constructs. A proprietor's right to exclude may come from the fact that they are in control of the software code which runs the platform, but that power is reinforced through the rules we choose to apply to protect that code and the servers upon which it runs and the interpretation we give to the contracts which condition access to the platform. At each of these stages we are embedding certain values in the legal system, and it is wrong to suggest that they form part of a natural state of affairs in which the government ought not interfere. There are clearly choices to be made.

Fundamentally, our legal system should develop to arrive at just results. This means that we must be careful to consider the interests of participants and refrain from marginalising those against the interests of the proprietors. This is a balancing exercise, however, because we must be careful not to destroy the vibrant and diverse nature of these spaces through over-regulation, and equally careful not to undermine the commercial viability of platforms which are provided through private sector investment.

This balancing process is the essence of digital constitutionalism. The rest of this thesis will consider how public values can be applied to the laws which govern virtual communities.