In Paul Gowder’s recent blog post, as well as in his new book, he argues that we should democratize, rather than dismantle or restructure, Big Tech platforms. Echoing cyberspace analogies popular 20 or so years ago, Gowder analogizes private technology platform companies to states, and argues that law and regulation must keep these companies “intact” while transforming them into democratic fora. If only law could channel localized individual knowledge back to corporate centers of decision-making and make sure that tech companies followed their long-term rather than short-term interests, then platforms would be well-positioned to serve the public interest.
Gowder is far from alone in framing the landscape of possibilities as a choice between “breaking up” and “democratizing” Big Tech. However, as I will argue in this post, this framing obscures more than it reveals, and is symptomatic of three widespread blind spots in digital platform governance policy and scholarship: (a) an overemphasis on solutionism and a corresponding failure to adequately describe the problems at hand, (b) an obsession with categorical tradeoffs and disciplinary siloes, and (c) an impoverished account of the political economy of technology, of the co-evolution of politics and production, and of the core role of material infrastructure in digital settings. After setting out these blind spots, I introduce a practice of regulatory experimentalism, which I see as the best way forward in technology platform markets.
Tech Regulation Is Complicated
There is a tendency among tech lawyers to assume that selecting the appropriate legal and policy solutions can precede a deep understanding of the large families of diffuse democratic, economic, and material threats that motivate intervention. Yet, too often, the methods and solutions that political science and law can offer are broadly inadequate to the questions and problems at hand. By reducing the question of tech regulation to a problem internal governance, Gowder caves to this urge to solve a broad and multifaceted problem with relatively simple tools.
Consider, for instance, the wide range of threats presented by Big Tech platforms such as Meta or Alphabet. They are replacing the public sphere with privately curated spaces for public interaction that are increasingly structured and driven by profit-maximization goals. They use their market advantage – in particular their control over data, as well as algorithmic and technical processes – to police and foreclose the opportunities of new entrants or less powerful suppliers in the digital economy. Most crucially, perhaps, decades of neoliberal governance have allowed these companies to create and broadly self-govern the technical and epistemic landscape within which commerce, democracy, and policymaking can take place.
The last point is key. Nowhere can the localized knowledge crucial to Gowder’s democratic vision be elicited in isolation from the power and influence that Big Tech platforms wield over knowledge-producing and value-creating infrastructures. The platform economy is organized around vertical individualized contractual consumer-to-platform or business-to-platform relations. These vertical relations allow platforms to architecturally limit the possibility of collective, horizontal relations between persons on the platform.
The 2018 Cambridge Analytica episode highlighted how various third parties work to profile individuals so as to personalize content, advertising, and the user-experience on behalf of political parties. It underscored that platforms’ infrastructural affordances contribute to social fragmentation and put the possibility of shared collective knowledge and of collective action at risk. It is nearly impossible to imagine users horizontally uniting and exerting counter-power against tech platforms’ content decisions, for example in Myanmar, in a fragmented landscape organized around individualized vertical relations. In other words, democratizing and distributing knowledge in the digital platform economy requires questioning the existing power and control tech companies hold over the infrastructures through which knowledge is produced and value is created.
Second, and related, legal scholars and policymakers like to posit abstract trade-offs between pre-existing domains of law, such as antitrust, utility regulation, and privacy, which they treat as distinct silos. Yet rarely are these domains at odds with one another. Over the past 50 years, there have been several scholarly and judicial attempts to paint antitrust as a deregulatory domain of law at odds with lasting forms of industry regulation. These efforts have partly influenced the now widely accepted status of antitrust law as a residual substitute for sector-specific regulatory frameworks. In Trinko, for example, Justice Scalia stated that where a regulatory structure designed to deter and remedy anticompetitive harm exists, “the additional benefit to competition provided by antitrust enforcement will tend to be small.” A serious approach to digital platform markets requires contesting this position and recognizing that antitrust enforcement, merger reform, and sectoral regulatory frameworks such as AICOA and the AMERICA Act are not mutually exclusive substitutes and instead need to co-exist. There is no reason to favor one over another, or to dismiss one on the ground that another is more promising.
Further, we should resist one-size-fits-all approaches to the digital economy. One reason for pursuing a flexible and ecumenical approach is that the very structure of the digital economy is still evolving: as Big Tech grows stronger and more politically and economically entrenched, the justifications for structural approaches aimed at decentralizing centers of economic and political decision-making grow more urgent. The legal context also matters: if the US had strong tech regulatory frameworks similar to those of the EU, arguments that antitrust should retreat in line with Trinko might gain greater plausibility. Finally, the ideological and political perception of Big Tech companies is also changing: the current trial against Google reveals a much stronger appetite for enforcement than used to be the case in 2013. Given this developing technological and legal landscape, we should avoid advocating myopic or atemporal solutions that are maladaptive to infrastructural, legal, and ideological change.
Finally, Gowder’s approach is part of a broader genre of liberal legal scholarship that emphasizes “democracy” to the detriment of “political economy.” In the US, much of the early work on platforms focused on self-moderation without worrying much about the economic rationales that motivate platform companies to exist and operate as they do. In addition, decades of libertarian and pro-business speech jurisprudence have led legal scholars to be generally reluctant to propose regulating infrastructure in the digital platform economy. The result is an impoverished landscape of proposals that prefer to keep tech companies “intact.”
But what does keeping tech companies intact mean? Taken seriously, democratization requires challenging the very channels of communication that tech platforms control, the structure through which knowledge is constructed and monetized, and spreads. A serious democratic approach is thus incompatible with “keeping tech companies intact.” It is impossible to democratize an economy organized around knowledge asymmetries, opacities, dependencies, and inequalities without advancing antitrust or other infra-structural action aimed at displacing tech companies’ asymmetric power over their users, customers, and competitors (see for example the DOJ’s cases against Google or the FTC’s complaint against Amazon).
Identifying these blind spots suggests that a richer approach to the governance of digital platforms and their corporate owners is possible.
My own approach in recent work has been to resist the urge to provide “solutions,” instead promoting a practice of regulatory imagination and experimentation. Such practice can be grounded in the regulatory experimentalism that accompanied the emergence of utilities and infrastructural regulation in the late 19th and early 20th century. It can also find inspiration in the LPE movement itself or in NPU law. Conceived as a practice, infrastructural experimentalism can guide policymakers and scholars of the digital economy toward more adaptive forms of legal and technological constraint.
As I see it, the practice must be centered around three ideas. First, regulatory experimentalism must learn from the legal and policy errors of the past. The failure to prevent increasing centralization in the face of the decentralizing and deregulatory impulses that characterized the rise of informational capitalism should be recognized in today’s discourses around cryptocurrencies or AI. Similarly, we must be ready to scrutinize how liberal individualism and the fetish around networked individual preferences, choices, and speech is contributing to new forms of authoritarianism.
Second, regulatory experimentalism must take infrastructure and markets seriously as the terrain of regulatory and emancipatory possibility. In the face of renewed efforts to enclose and privatize digital productive processes, taking control over our shared infrastructure and productive processes is necessary if we are to realize the possibility of collective self-determination. “Democratizing” digital platform environments therefore requires taking the residual infrastructural, material, and economic power out of the hands of digital platform companies and their shareholders. New proprietary structures and corporate arrangements could promote a more diversified digital economy that is not only more structurally decentralized and multi-polar, but also more diverse and infused with public values in its funding, functions and governance models.
Third, regulatory experimentalism requires a playful approach to the interaction between economic branches of law such as antitrust and utilities, with a focus on advancing the public interest. Such experimentation may or may not bring the short and long-term interests of private companies into alignment with the public interest. That’s because the self-interest of private companies ought not to be the primary constraint that shapes policy.
Focusing legal efforts on structuring markets and productive processes so that they are more representative is an essential part of democratizing digital platform ecosystems. Instead of taking “democracy” out of the market, the task is to structure the digital economy through law in ways that make it globally more responsive to public demands across platform and corporate boundaries.