Governable spaces: a feminist agenda for platform policy

The business models of the online economy have largely operated on the assumption that a single neutral platform, a universal set of rules, and standardised infrastructure can serve diverse users without prejudice (Gillespie, 2010; Milan & Treré, 2019). Experience seems to suggest, however, that this approach exacerbates existing inequities. Along lines of gender, for instance, female-presenting social media users face disproportionate hostility and exploitation (Lewis et al., 2017; Schoenebeck et al., 2021; SSL Nagbot, 2016), while gig platforms have reinforced gendered segregation and subjugation in labour markets (Arcy, 2016; Schor, 2020). Dominant policy responses have nevertheless embraced the platforms’ claims of being neutral infrastructures, conferring to platform companies both the power and responsibility to govern from the top down (Klonick, 2018; Plantin & Seta, 2019). This regime appears to be facing a mounting crisis, as users increasingly perceive its determinations as arbitrary (Myers West, 2018; Nurik, 2019).

I propose here to extrapolate from feminist insights to formulate an alternative policy agenda based on cultivating governable spaces. I apply this agenda to vexing challenges in three domains of the online economy: social media communities, platform-mediated work, and network infrastructure.

A recurrent theme in feminist critiques of internet culture, and feminist thought more generally, is suspicion toward allegedly neutral, open, non-political forms of organisation (Freeman, 1972; SSL Nagbot, 2016). Feminists have extended earlier analyses of under-valued labour such as housework (Federici, 2012) to emerging manifestations of informal digital labour (Cady, 2013; Jarrett, 2014; Nakamura, 2015; Rankin, 2018; Schneider, 2021c). Critiques of facial recognition (Buolamwini & Gebru, 2018), surveillance (Browne, 2015), and utilities such as search (Noble, 2018) demonstrate that services designed for “anyone” may in fact do harm to people whose bodies and experiences are not those of the designers and investors. Feminist tradition sees with particular clarity how the online economy has contrived to be both apparently open and persistently unaccountable. I argue that feminist tradition also presents fitting response: systems that organise our technology should be governable by the people who rely on them.

The dominant patterns of the online economy bear some resemblance to what activist-scholar Jo Freeman observed on smaller scales among feminist “rap groups” in her famous essay against “The Tyranny of Structurelessness” (Cohen, 2018; Freeman, 1972). She found that groups lacking clear processes or pathways for participation often fall into rigid hierarchies and resentment. “Those who do not know the rules and are not chosen for initiation must remain in confusion,” she wrote, “or suffer from paranoid delusions that something is happening of which they are not quite aware” (p. 152). In response, she offered proposals for “democratic structuring” among activist groups. More recent intersectional feminists have similarly seen fit to establish intentionally bounded gathering spaces, along lines of gender identity, race, and affinity (Savic & Wuschitz, 2018; Toupin, 2014). Within such spaces, participants have found the capacity to establish protective codes of conduct and clear lines of accountability (Finley, 2018). Feminist media studies have elevated participation and community control as a strategy for people seeking to transform media environments that have historically marginalised them (Kember & Zylinska, 2012; Zobl & Drüeke, 2012). When a Malaysia-based international process developed “Feminist Principles of the Internet”, the principle of “governance” extrapolated this logic with a call to “democratise policy making affecting the internet as well as diffuse ownership of and power in global and local networks” (Association for Progressive Communications, 2014, n.p.). In feminist practice, I contend, lies the foundation for a global policy strategy.

The pursuit of governable spaces is an agenda for technical and legal platform policies that take this strain of feminist insights seriously. Governable spaces are systems that enable participants to deliberate, make decisions, and enforce them through accessible, transparent processes. Achieving governable spaces in this sense requires bringing democratic structuring into sites of domination and structurelesseness.

The feminism considered here includes a range of sources, not necessarily ones belonging to a single wave, strand, or lineage. Surely there are feminists who would challenge the tendencies I highlight. I am not seeking to alter or summarise feminist tradition, but to identify patterns that it has seen especially clearly. I use Freeman’s essay as a centre of gravity that attracts shared concerns among diverse feminist perspectives. These reveal ways in which male-dominated technology companies have leveraged patriarchal relations into structures of top-down control. Patriarchy thus reconfigures itself as an allegedly benevolent paternalism through the “exertion of positive rather than coercive power”, as Liena Gurevich (2008) puts it. The prevailing discourse on internet platforms tends to take such paternalist rule for granted as necessary and just; feminist tradition has far less patience for doing so.

This essay extrapolates from feminist scholars’ attention to self-governance, outlining an agenda for governable spaces across three domains of controversy and policy-making in the online economy.

First, in the context of social media communities, governable spaces should empower users to shape the decisions that affect them, such as policies about acceptable content and the terms of participation. Second, in the realm of “gig economy” labour platforms, I highlight a proposal for a worker-managed marketplace, grounded in the experience of care workers. Finally, at the level of underlying infrastructure, governable spaces can demystify what are usually invisible operations—transforming networks’ substrates from the preserve of distant experts into an accessible commons. Together, these proposals speak to technologists and activists who might test new strategies, as well as policymakers who might eventually enshrine them in the expectations of law.

The upshot of each exercise is to reconfigure supposedly neutral regimes, frequently managed through private monopolies, with architectures of self-governance and accountability. Doing so involves new forms of boundary-making and rule-setting, against the ambitions of corporate monocultures. The “agenda” I refer to seeks not rigid central planning but lived environments crafted collectively over time, as we inhabit them.

For communities

During the waning days of 2020, US president Donald Trump threatened to veto the annual National Defense Authorization Act if it did not include a provision unrelated to national defence: a reversal of Section 230 of the 1996 Communications Decency Act. This was one more instance of melodrama in the career of a snippet of law known, in the title of a book-length study on it, as “the twenty-six words that created the Internet” (Kosseff, 2019). For speech that occurs on an “interactive computer service,” it pins liability not on the service but on the user. Critics have accused Section 230 of both absolving major social-media firms for whatever filth users happen to post and for facilitating censorship.

Despite its reputation as a permission slip for online disorder, Section 230 cleared the way for a new order of governance. This order spread far beyond the United States with the global influence of Silicon Valley platforms and parallel regulations in other jurisdictions. The legislators who wrote the provision hoped their safe harbour would encourage services on the fledgling internet to self-regulate their content without fear of political meddling (Kosseff, 2019, p. 248). The result has turned platform companies into what Kate Klonick (2018) has called “the new governors”—not merely enforcers or managers, but architects of meaningfully autonomous governance regimes over users and the public sphere. Internationally, the assortment of governors is becoming ever more complex. The internet appears to be drifting toward a “splinternet” of conflicting regimes, requiring global platforms to behave differently in different jurisdictions, if they are allowed to operate at all (Lemley, 2020).

From the perspective of most social media users, platform governance is a matter of imposition, whether by remote company owners or by the more proximate volunteer administrators and moderators. The design practice of “implicit feudalism” (Schneider, 2021a) relies on power-holders who are not chosen or removed by those they govern. Rule enforcement occurs through censorship of user content or the removal of users altogether, but rules do not necessarily apply to the administrators themselves. Users can speak out or leave online spaces, but they lack direct procedural power (Frey & Schneider, 2021). This regime plausibly contributes to the “techlash” against platform companies that spreads with every scandal of content moderation and abuse; by hoarding power, the companies have hoarded the blame.

Feminist political theorists have retrieved and radicalized Aristotle’s recognition that healthy governance grows from the ground of friendship among citizens (Devere, 2013; Schwarzenbach, 2005). Marilyn Friedman (1989) argues for centering peer relationships, rather than the patriarchy-inflected family or territorial community, as the starting point for liberatory politics. Even while social-media platforms have opened new opportunities for friendships among peers, instrumental power flows from the top down, from company CEOs down to the feudal admins and mods, according to terms that government regulators set or fail to set.

Among the more measured proposals for reforming Section 230 in the United States is a call for removing protection from platforms that act as “bad Samaritans” by actively encouraging toxic or criminal content (Citron & Wittes, 2017). But this proposal presumes a platform-centric approach to content policy-making, rather than one centred on the policies communities might make for themselves. The company-customer relationship so central to industrial markets remains the preferred logic of regulation, not the peer-to-peer relations that prevail in so much of online life. For instance, the European Union’s General Data Protection Regulation places obligations on companies and establishes rights for individuals such as the “right of erasure.” Agency has been placed largely in the hands of company bureaucracies and zealous users, while the potential for collective action is only beginning to be explored (Casarosa, 2020). Yet individual users have only rarely been able to exercise their rights under the law; the essentially collective nature of platforms’ data collection practices means that meaningful enforcement will depend on the kinds of voluntary associations among peers that Friedman turns to, the building blocks of feminist politics.

What if legal protections for platform companies hinged on fostering governable spaces, in which the rules and their enforcement were accountable to users in clear and transparent ways?

One step toward making community spaces more governable is for users to to establish clearer boundaries and purposes where they gather—echoing Virginia Woolf, spaces “of our own” (Fiesler et al., 2016; Toupin, 2014). These might look like the independent, interconnected “instances” of a federated social network (Zignani et al., 2018), or like the semi-autonomous corners of a centralised platform like Reddit (Fiesler et al., 2018). Such scaled-down autonomy seems to work, even against highly organised toxicity. In the “fediverse,” feats of coordinated sanctions have successfully isolated instances such as those used by the Islamic State and the far-right network Gab (Caelin, 2020).

Size or boundedness alone, however, do not make a space governable. Governable spaces require tools that support a variety of participatory mechanisms for rule-making and administration, such as elections, petitions, boards, and juries (Frey & Schneider, 2021; Schneider et al., 2021; Zhang et al., 2020). Rather than relying on assignments of permissions to individual users, default settings might assume decision-making as a collective affair. For instance, the European food-sharing platform Karrot allows a local community to remove a member only through a group process, rather than by the fiat of a single administrator (djahnie, 2019). While such an arrangement lies well outside the norms of social-media systems, it is at home in feminist conceptions of the relational self, the person as a nexus of relationships (Friedman, 1997/2018).

Policy-makers might push such experiments into the mainstream through mechanisms of corporate accountability. Some of the largest platforms have already begun to dip their toes into this water voluntarily, such as through Facebook’s formation of an Oversight Board to review content-moderation decisions (douek, 2019–2020) and Airbnb’s advisory board for user-hosts (Airbnb, 2020). These nascent corporate policies remain severely constrained in their power and independence from management, but that could change. Governments might impose a variant of Germany’s requirement of worker co-determination through participation on corporate boards (Jäger et al., 2021); in this case, platforms of a certain size might need to have user-elected representatives on their boards or moderation teams.

The remedy for widespread toxicity need not lie solely in setting policies from above. If social platforms became more governable kinds of spaces, the responsibility for what happens on them could be more sensibly shared.

For workers

Before she was a legal scholar, Sanjukta Paul worked for a labour union. She saw how US antitrust laws in particular narrowed the options workers had for joining together and organising, undermining worker power. Policy, she came to realise, acts as an “allocator of coordination rights”: an arbiter of who is allowed to team up and how (Paul, 2020). While the constraints of US antitrust doctrine on labour organising are specific to the country and context, law everywhere plays this allocative role.

The absence of coordination can be difficult to notice, lurking in the shadows of what the law prohibits. Feminist scholars have chronicled how the policing of coordination has been used to undermine women’s collective power historically, such as through the persecution of witchcraft in early-modern Europe and America (Federici, 2004). Much as witch hunts sought to keep women’s economic lives dependent on patriarchal systems, laws today help preserve a fragmented, atomized workforce available for gig platforms and other precarious jobs. While antitrust law is only rarely wielded against large platform companies, in many countries it currently imposes legal barriers that have prevented platform workers from forming unions (Johnston & Land-Kazlauskas, 2018). Paul invites us to ask: Who is and isn’t allowed to find common cause?

Economist Juliet Schor’s After the Gig (2020) troubles the corporate story of a liberating, platform-mediated “gig economy”, without dismissing it altogether. For decades, her work has illustrated the illusions in economic common sense, often through attention to under-studied job-sites such as housework and childhood. After the Gig draws on close-up studies of platform workers—the drivers, the deliverers, the hosts, the doers of various tasks—and reveals their cleavages. Some workers do find a kind of liberation, while others fall into a trap. There is no one gig or sharing economy, but many distinct and often isolated experiences.

Schor and her research team constructed a kind of conversation that is otherwise missing across a field of dispersed experience. Unlike social-media platforms, gig apps discourage persistent lateral relationships among users, whether they hold worker or customer roles. The platform claims to supplant the need for relationships. Such user-experience designs, like early-twentieth-century US antitrust laws, establish policies of coordination rights. The platforms organise those rights on behalf of managerial control, of a kind tighter than on any industrial shop-floor. Wherever platform companies have succeeded in suppressing unionisation, users get open and frictionless transactions but no durable means of seeing each other, of comparing experiences, of finding the wherewithal to co-govern.

Feminists have long sought to reveal and recognize the significance of work that patriarchy would prefer to keep invisible. From unwaged housework to underpaid roles in offices and factories, “feminised” jobs have been systematically undervalued (Federici, 2020; Jarrett, 2014). Before online gig platforms, women performed piecework for the textile industry under similarly precarious regimes; women’s critical roles in building the computer industry have suffered persistent neglect (Rankin, 2018; SSL Nagbot, 2016). Thanks to its attention to this history, feminist tradition brings conceptual tools that speak to the present abuses often euphemized as “the future of work.”

Feminist scholars have identified how emotional labour and reproductive labour enable the digital economy to function, while the credit for production typically goes elsewhere (Arcy, 2016; Jarrett, 2014; SSL Nagbot, 2016; Terranova, 2000). Tech companies increasingly depend on little-seen and poorly rewarded “ghost work” that occurs in homes or offices far from the tech hubs (Gray & Suri, 2019). If a social media company succeeds in removing violent imagery from its platform, is that because of the executives’ policies and the engineers’ algorithms, or the offshore workers who have to look at things all day they will never be able to un-see? Workers’ unpaid family members may organise meals and schedules that make the paid work possible. Acts of governance happen not just in executive boardrooms or shareholder votes, not just in algorithms and user-experience, but in the daily negotiations that companies intentionally hide from view.

Schor remains hopeful about the possibility of a freer, less wasteful, platform-mediated future of work. But “achieving the potential of platforms requires specific conditions,” she writes. “They won’t be met if today’s corporate elites are in control” (2020, p. 2).

The mode of transferring control that Schor considers most closely is that of cooperative platforms, owned and governed by their workers or users (Scholz & Schneider, 2016). Her primary case study, the Canadian company Stocksy United, represents a small but vibrant slice of the online stock-photography market. Stocksy’s artist-members, located in dozens of countries, can discuss platform policies through an online forum and vote on major decisions. The prospect of such an artist-centred platform has enabled the co-op to thrive and attract skilled creators away from much larger, investor-owned competitors.

Financial-system reforms could enable these kinds of governable spaces to become more widespread, if groups of people with common interests had access to capital now available only to wealthy investors and funds (Pentzien, 2020; Schneider, 2021b). If democratically owned companies ran platforms of all sorts, surely they would be more likely to foster democratic spaces for their users. User-governable companies can begin with creative entrepreneurship but will require public policy to achieve meaningful scale.

Ra Criscitiello, the Deputy Director of Research for Service Employees International Union-United Healthcare Workers West (SEIU-UHW) in California, attempted to build a platform co-op with her union’s members. NursesCan, as they called it, was to be a gig platform for at-home, on demand healthcare services. But building a viable co-op in a tech economy made for investor ownership and venture capital did not prove feasible. Criscitiello responded by becoming more ambitious.

In the storm of California’s policy struggles over gig work, she initiated a state-level proposal called the Cooperative Economy Act, a version of which has since been introduced before the state legislature (Gonzalez, 2021). The bill creates a federation of tax-advantaged, employee-owned cooperatives that could contract with online labour platforms. These would enable workers to collectively bargain over the terms of their work for platforms, without the platforms needing to employ the workers directly. Workers would participate in electing their co-ops’ leaders. Although California’s 2020 law known as Proposition 22 exempted platform drivers and delivery workers from the rights associated with employment, other platform workers—such as SEIU-UHW’s nurses—stand to benefit from organising their gigs as a collective. The proposal shares features with the Hollywood system, devised long before the internet, which enables the film production workforce to move from job to job while retaining union representation and even profit-sharing with the studios (Fisk, 2018).

Criscitiello’s proposal is reminiscent of another earlier effort to put care work under democratic control: Italy’s social cooperatives. These cooperatives first arose as the Italian state withdrew from direct provision of care services, such as for the elderly and disabled (Thomas, 2004). Unlike the state providers, social cooperatives are frequently co-governed by their workers and beneficiaries, together. The model grew to hundreds of businesses, and in 1991 the national government established social cooperatives in law, providing tax treatment more appropriate to the social benefits they provide. Thousands more formed. Statutes for social cooperatives have since spread to other countries in Europe and beyond. At the same time, a new kind of cooperative has been emerging that offers shared services for freelance workers (Martinelli et al., 2019). The pioneers include Smart, which now operates across most of Europe, and Italy’s Doc Servizi. Like Stocksy United, both first emerged among artists.

Even in the absence of legal cooperatives or unions, workers are finding ways to gain fuller control over their livelihoods. Luis H. Lozano-Paredes (2021), for instance, describes the informal self-organisation among platform-based drivers in Colombian cities, who have lessened their dependence on corporate ride-sharing platforms with technologies they more fully control. They have developed guild-like clubs with exacting rules for membership and conduct, while handling payments through peer-to-peer apps. Whether or not governments recognize their entrepreneurship as legitimate and welcome, workers like these are surviving by governing spaces of their own.

For infrastructures

Magnolia Road Internet Cooperative is made up of neighbours who provide internet service for each other, spanning a poorly connected stretch of Rocky Mountain foothills in the US state of Colorado (Schneider, 2016). The co-op’s closet-sized locker, rented at a local storage facility, holds a mess of routers, wires, and antennas. Consumer-members of the co-op learn to install, use, and repair the equipment themselves. According to the way many people are taught to think about internet access, this does not seem possible—surely such matters are comprehensible only to the kinds of national telecom monopolies that have to be paid to bestow connectivity.

Latin America has a long legacy of microtelcos that provide phone and Internet service in communities that corporate providers do not see fit to serve (Galperin & Bar, 2006). Community radio, another form of information infrastructure widespread in Latin America, has enabled women to make their voices heard outside traditional gender roles; that practice also went online with the early internet, such as with the Feminist International Radio Endeavor (FIRE) in Costa Rica (Arriola, 1992/2001; Thompson et al., 2005). As Paula Serafini (2019) argues, in a study of Argentinean cases, community radio stations have served as an ecofeminist “space of care” for communities engaged in ongoing resistance to extractive economies and cultures.

Feminists have challenged how patriarchy mystifies science and technology, casting them as domains beyond the possibility of comprehension for all but a few (Harding, 1986; Wajcman, 1991). Mystification obscures the economics of accumulation behind new technology, turning our attention to the innovation instead of the extraction it enables (Russ, 1978; Supp-Montgomerie, 2021). As Lisa Parks (2007) has shown, utility firms construct infrastructure so as to be not only incomprehensible but invisible—underground, overhead, or disguised as natural phenomena like trees or rocks. The task of comprehension has required interventions like artist Ingrid Burrington’s unofficial guidebook to the Networks of New York (2016). What users cannot understand or notice, users cannot govern.

Infrastructure dictates what people have available to them, on what terms, and at what cost (Edwards et al., 2007). It requires labour, often carefully hidden from view, to produce and maintain (Parks, 2015). Corporations do not typically build infrastructure with the intent of enabling users to govern its operation. This sort of hardware is supposed to be the purview of remote experts, out of ordinary users’ reach. Yet making infrastructures governable has demonstrated the capacity to address the market failures of private interests. It was only when the US government began financing electric cooperatives in the 1930s that most rural areas in the country got power lines (Oakland, 2020). Evidence suggests that cooperative and municipal broadband systems have advantages of cost and quality over corporate control (Ali, 2021; Talbot et al., 2017). Community-based connectivity efforts in many contexts—from urban Detroit (detroitcommunitytech.org) to towns across Catalonia (guifi.net)—show that user-governed networks can succeed where investor ownership falls short.

Governable infrastructure requires an inversion of prevailing policy regimes. In some cases, including many US states, jurisdictions have acceded to laws that prohibit municipal or cooperative broadband services from competing with private firms (Talbot et al., 2017). But even where shared ownership is an option, it frequently lacks the access to capital necessary for infrastructure investments. The current pattern of prohibitions could instead be reversed; jurisdictions might ban long-term investor ownership of physical internet infrastructure. Private firms might build and help capitalise such projects, but the law could ensure that communities of users must become the stewards after the build-out. Doing this would require a public commitment to financing access in under-served areas, but as the legacy of US rural electrification suggests, such investments are well worth the cost.

Software can also serve as infrastructure, particularly the protocols and platforms that large sections of an economy rely on (Plantin et al., 2018). Governments can support the development of governable platforms by adopting procurement preferences for commons-based software projects, such as the German government’s use of NextCloud for collaborative file-sharing (Nextcloud, 2018) and France’s adoption of the Matrix chat protocol (Hodgson, 2018). That same software can then be deployed and self-managed by communities anywhere in the world. For instance, the explicitly feminist cooperative in Barcelona FemProcomuns (femprocomuns.coop) uses NextCloud as part of its “Commons Cloud” platform. Investments in tools like these enable people to move more of their digital lives into governable technology stacks (Schneider, 2022).

For both software and hardware infrastructures, public incentives need not assume perpetual investor control over essential services. Rather, policy could encourage an expectation of “exit to community” (Mannan & Schneider, 2021), where the result of successful entrepreneurship is a transition to community control. The patriarchal metaphors of “market dominance” and “liquidity event” that populate the jargon of startups (Brandel et al., 2016) guide them toward investor-focused incentives and are inadequate for producing governable, sustainable digital infrastructure. Policy should ensure that building governable spaces is a real option for technology entrepreneurs.

Conclusion: Platform paternalism

The aspiration of fostering governable spaces on and around digital networks may evoke concern for some. Marginalised communities have tended to receive the brunt of hate speech and abuse online (Schoenebeck et al., 2021). Must these same people now also take on the extra labour of self-governance? Meta CEO Mark Zuckerberg has been especially explicit in defending his company’s size and power on the basis that scale is necessary to support the costs of protecting users from each other (Feiner & Rodriguez, 2019). As the complex of abuses grows, Silicon Valley CEOs such as Zuckerberg—who tend not to share the experience of marginalised identities themselves—insist that they alone can solve these problems of their own making. They contend that their spaces are open, are in principle accessible to all, with neutral standards applied to all. Why shouldn’t people accept their offer, however paternalistic?

There is mounting evidence against the efficacy of large-scale, platform-centric content moderation compared to community-level approaches (Myers West, 2018; Seering, 2020). Yet community governance has hardly been tried. Users of online platforms have had too little chance to make decisions on moderation policies, agree on basic rights for workers, or build their own infrastructures that investors have failed to provide. The longer users lack these opportunities, and the resources to utilise them, the more tolerant they become of paternalistic control. With even a fraction of the investments companies now make for protecting and perfecting top-down control, platforms could explore strategies that make co-governance sustainable rather than burdensome. But the strongest argument for governable spaces is essentially a priori: Are we really willing to accept anything else?

The agenda I have outlined, motivated by feminist insights on crafting intentionally governable spaces, confronts multiple assumptions that tend to prevail in the reigning, paternalistic online spaces. First, platforms typically teach their users to expect that when problems arise, the platform companies can and should be trusted to address them, with minimal transparency and accountability. A premise for this message, further, is that social problems are best understood as engineering problems—making the engineer-rich platform companies uniquely suited for finding solutions. Finally, crises of accountability on platforms have prompted activists and company executives alike to ask for both external government regulation and decisive action from the CEOs. But other kinds of solutions can begin with empowering user communities to design and carry out reforms for themselves.

I have sought to outline an agenda that breaks from paternalist habits. As long as daily online life remains mired in structurelessness, top-down control will seem like better than nothing. But if governable spaces are a live option, the paternalist promises will reveal themselves for what they are.

Acknowledgments

The author is grateful for feedback on earlier drafts from Ra Criscitiello, Angela Daly, Frédéric Dubois, Esperanza Miyake, Olu Niyi-Awosusi, Blake E. Reid, Emily van der Nagel, and Amy Zhang.