Platform cooperativism and freedom as non-domination in the gig economy

Abstract

While the challenges workers face in the gig economy are now well-known, reflections on emancipatory solutions in political philosophy are still underdeveloped. Some have pleaded for enhancing workers’ bargaining power through unionisation; others for enhancing exit options in the labour market. Both strategies, however, come with unintended side-effects and do not exhaust the full potential for worker self-government present in the digital gig economy. Using the republican theory of freedom as non-domination, I argue that G.D.H. Cole's 20th-century defence of guild socialism offers a promising avenue for enhancing worker autonomy in the gig economy. Platform companies rely on relational and structural domination to undermine worker autonomy, but Cole's guild socialism was specifically designed to enhance autonomy in the workplace. Following Cole's advice today would amount to a defence of worker-owned cooperative platforms. By putting workers in control of platform design and governance, cooperative platforms create new opportunities for worker autonomy. However, platform cooperativism faces serious challenges if it plans on becoming a realistic alternative to the traditional gig economy.

Take this for the last word of my dream of what is to be: the test of our being fools no longer will be that we shall no longer have masters.

– William Morris

Introduction

In 2022, food-delivery platform Deliveroo announced its departure from the Netherlands. The official motivation was a decline in consumer demand due to inflated food prices, but a court ruling had also just reclassified Deliveroo couriers as employees rather than independent contractors. This decision dampened Deliveroo's financial prospects, so the company halted its operations entirely in the Netherlands. This outcome threatened to push gig workers into unemployment. Full-time gig workers end up on labour platforms because they find few viable alternatives on the formal labour market. The gig economy sometimes acts as an employer of last resort. However, the courier Martijn Muller and some colleagues used this opportunity to launch a cooperative platform of their own called ‘BestellenBij’. This cooperative could absorb the now unemployed Deliveroo couriers and establish more democratic worker self-government with fairer remuneration rules. Compare this crisis averted to another Dutch case from a few months later. In January 2023, the domestic work platform Helpling filed for bankruptcy after another reclassification lawsuit in which Helpling workers were redesignated as temp workers rather than self-employed entrepreneurs. FNV, the largest Dutch trade-union, welcomed the court decision and subsequent bankruptcy claiming that companies unable to properly pay their workers ought not to exist (De Ruiter, 2023). Labour market experts were, however, less enthusiastic (Arets, 2023; Christiaens, 2023b): domestic workers would be left without a job or would be forced into the informal shadow economy, where they would be even more exposed to exploitation or sexual harassment. Such outcomes could have been avoided if only there had been a more ethical alternative like BestellenBij.

In this paper, I provide a political-philosophical defence of platform cooperatives as such an alternative. Most literature focuses on the injustices of the gig economy (Cant, 2020; Kessler, 2018; Ravenelle, 2019; Rosenblat, 2018; Woodcock and Graham, 2020), but reflections on emancipatory alternatives are less prolific. In political philosophy, most attention is devoted to union organisation and collective bargaining (Hickson, 2023; Rogers, 2023). But, as the Helpling example shows, union activism alone might be insufficient and can generate unintended side-effects if the platform company disengages from bargaining compromises. Moreover, collective bargaining grants workers some ‘antipower’ to resist corporate decision-making (Pettit, 1996), but not genuine self-determination (Cole, 2017a: 19). This strategy does not fully explore the range of available emancipatory practices. As Robert Ovetz writes, ‘even if [gig workers] become employees or negotiate higher wages the app is still left firmly in charge’ (Ovetz, 2023: 282). Another option is to enhance workers’ exit options through, for example, a job guarantee (Bieber, 2022), which would indirectly enhance the collective bargaining power of gig workers as well. If the state guarantees more jobs on the labour market, labour platforms would experience more difficulty convincing workers to accept their job offers. However, this measure would still involve what Nathan Schneider has called the ‘implicit feudalism’ of platform governance (Schneider, 2022). Platform companies would still unilaterally set the terms and conditions, while workers would only have the freedom not to accept. Their voice would still depend on the goodwill of platform companies allowing their input to be heard. Genuine self-government is not part of this project.

Platform cooperativism, on the other hand, has received little attention in political philosophy, yet it has had a large impact in activist circles. ¹ In 2014, the Platform Cooperativism Consortium was founded and, currently, there are worker cooperatives across the gig economy, like CoopCycle, Up&Go, Stocksy United, and Drivers Cooperative, to name just a few. In contrast to the previous solutions, platform cooperativism specifically emphasises the need for worker self-government. I use G.D.H. Cole's republican philosophy of guild socialism to defend this strategy. Cole is an early 20th-century socialist philosopher linked to the UK Labour Party, but he was most famous for identifying not poverty or exploitation but domination as the foundational wrong of capitalism: ‘Poverty is the symptom; slavery the disease. The extremes of riches and destitution follow inevitably upon the extremes of license and bondage. The many are not enslaved because they are poor, they are poor because they are enslaved’ (Cole, 2017b: 111). This socialist-republican concern for freedom as liberation from wage-slavery makes Cole an interesting source for articulating a political philosophy of the gig economy (Muldoon, 2022b: 81). After delineating in the first section how the conundrum of domination and wage slavery plays out in the gig economy, I devote the second section to outlining Cole's guild socialism and its applicability to worker cooperatives in the platform economy. In the third and final section, I respond to three common criticisms of platform cooperativism: (a) the concern that worker cooperatives would overcharge consumers for suboptimal services, (b) the Marxist criticism that cooperatives inevitably succumb to competitive pressure in a capitalist economy, and (c) the prediction that cooperatives tend to create new, implicit hierarchies.

The problem of domination in the gig economy

The major labour platforms in the gig economy (Uber, Lyft, Deliveroo, etc.) emerged in the 2010s with a business model that radicalises outsourcing techniques from the 1990s (Christiaens, 2022: 25; Degryse, 2023: 107). Multinational corporations are advised to directly employ only a small workforce devoted to ‘core competences’, while outsourcing other tasks to a periphery of precariously employed workers. This flexible layer of subsidiaries is hired through temp contracts, unpaid traineeships, franchise agreements and so on. At a university, for example, this means hiring a core group of professors and administrative staff, while outsourcing catering, cleaning and large parts of teaching and research to subsidiary companies, temp workers or externally funded researchers. Labour platforms have pushed this tendency to its extreme: they only hire data scientists and marketing specialists to maintain and promote the platform itself, while classifying actual workers as independent contractors. According to platform companies, gig workers are self-employed entrepreneurs who happen to use their platforms to meet customers (Rosenblat, 2018: 31).

This framework allows labour platforms to shirk traditional employer responsibilities and offload business risks onto workers. They cultivate a flexible workforce that can instantly expand or shrink depending on market demand, without the need to officially fire anyone (Dubal, 2017: 95). Uber drivers, for instance, can log in to the app whenever they like, but they only receive income for the actual tasks they perform. The risk of lost revenue during periods of low demand is borne not by employers, as in traditional labour contracts, but by workers themselves (Franke et al., 2023). The precarisation of gig workers also helps labour platforms build a ‘liability shield’ (Rogers, 2023: 110) against the legal responsibilities of traditional employers (Ravenelle, 2019: 37; Srnicek, 2017: 83). When workplace accidents happen, for example, platform companies deny legal responsibilities because they are not officially employing these workers (Christiaens, 2020; Gregory, 2021). Labour platforms, in other words, extract revenues from every task performed but transfer the responsibility of securing liveable incomes, social protections or workplace safety to workers.

From the perspective of republican political theory, this management strategy relies on relations of domination. Part of the republican tradition has already focused on workplace domination (Anderson, 2015; Claassen and Herzog, 2021; Muldoon, 2022a; O'Shea, 2020) or domination from digital platforms (Aytac, 2022; Hoeksema, 2023), but recent publications have focused more specifically on domination in the gig economy (Hickson, 2023; Muldoon and Raekstad, 2023; Zietlow, 2020). The core republican assumption is that

Your freedom as a person requires more than just being let alone, just benefitting from noninterference […] To be a free person you must have the capacity to make certain central choices – choices about what religion to practice, whether to speak your mind, who to associate with, and so on – without having to seek the permission of another. You must be able to exercise such basic or fundamental liberties, as they are usually called, without having to answer to any master or dominus in your life. (Pettit, 2014: xiv-xv)

Pettit identifies freedom with structural independence; the capacity to determine one's own actions without subjection to the uncontrolled will of another (Pettit, 2012: 50). Domination is its mirror image, the subjection of one's conduct to an alien will. Labour republicanism has specified that labour relations enact ‘private government’ (Anderson, 2017): workers are subjected to the one-sided dirigisme of their boss with few resources to enact genuine self-determination (Gourevitch, 2014; O'Shea, 2019; Vrousalis, 2021a). The gig economy is a poignant example of private government as gig workers often depend for subsistence income on access to labour platforms, yet have few tools available to determine the modalities of their conduct at work. Algorithmic decision-making often unilaterally stipulates which tasks are allocated to which workers, at what price and with which deadlines. If the only alternative to the algorithmic dominus is not to log off and forego necessary income, then gig workers cannot but work at the mercy of an alien will. They have to accept unfair policies or erratic policy changes every time they accept the app's terms and conditions (Ravenelle, 2019: 78). Trebor Scholz (2017: 57) has likened this choice architecture to Zugzwang, a chess term that connotes moments when a player is forced to make a move yet has no choice but to incur a loss. While nominally free, the player's options are forced. Two dimensions of domination can be distinguished: the relational domination from the labour platform as quasi-employer and the structural domination of economic precarity in a monopsonist labour market.

The primary dimension of domination in the gig economy is the direct subordination of workers to the platform company, mediated by the platform's digital infrastructure. Workers cannot actively determine their own conduct because they are beholden to decision-making algorithms that determine task allocation, payment rates, deadlines, etc. The platform company develops techniques of ‘algorithmic domination’ (Muldoon and Raekstad, 2023) or ‘digital Taylorism’ (Altenried, 2022) to impose its propensity for revenue maximisation on workers. Technology here functions as a conduit for the domination from platform companies (Delfanti, 2021b: 36; Papadimitropoulos, 2021: 248). ² Taylorist algorithmic management fosters a cognitive dependency of workers on the algorithms that coordinate their labour (Delfanti, 2021a: 43; Ovetz, 2023: 282; Savolainen and Ruckenstein, 2022: 2). As Braverman observed in his classic study of Frederick Taylor's scientific management, Taylorism focuses on separating the moments of conception and execution in the labour process (Braverman, 1975: 78–79). While management concentrates within its own ranks the power to conceive and coordinate the labour process and its products, workers are reduced to obediently executing managerial decisions. Workers are dispossessed of their knowledge of the labour process so that they become dependent on centralised workplace coordination to perform their labour.

Labour platforms similarly separate conception from execution by concentrating the coordination of the labour process in algorithmic decision-making processes, while workers are expected to execute algorithmic orders with few tools to question or adapt these commands to their own understanding of the labour process (Rogers, 2023: 61). Deliveroo couriers cannot know which orders to complete and how much revenue they garner without first submitting to Deliveroo's terms and conditions. Even then, couriers only know what the platform communicates to them. Delfanti (2021a: 41) documents a case of Amazon warehouse workers who have lost oversight over where all packages are stored, because only Amazon's AI-powered storage software controls the circulation of packages throughout the warehouse. Workers can do nothing but execute the orders appearing on their hand-held scanning devices. They become so dependent on the AI-system to perform their job that when this system disapproves of their work pace and subsequently fires them, they have little power or knowledge to contest the AI's judgment.

Once workers’ cognitive dependency on centrally organised labour coordination is established, labour platforms leverage this dependency to increase their own revenues at workers’ expense. The app is designed to maximise corporate income at the expense of workers, but this remains hidden behind the seemingly politically neutral platform design (Pasquale, 2020: 21). One example is the practice of dynamic pricing or algorithmic wage discrimination (Casilli, 2019: 132; Dubal, 2023a, 2023b). Labour platforms amass data about the pay rates of all accepted and declined tasks for each individual gig worker. Afterwards, they use this database to predict and offer the lowest acceptable remunerations. If the algorithm notices that some workers are so desperate for money that they will accept any price, it calculates exactly how low this price can go before workers log off. Different workers will hence get different payments for the same tasks, and vulnerable workers without the opportunity to decline tasks receive progressively worse wages for the same work. Because algorithmic rate-setting is hidden from view, workers have little knowledge or power to contest the platform's payment scheme. Most labour platforms unilaterally set pay rates without offering opportunities for questioning or critiquing these algorithmic decisions (Woodcock and Graham, 2020: 65).

There is not only direct, relational domination involved when workers are subjected to the arbitrary and uncontrolled power of platform companies, but also structural domination due to the socio-economic background conditions that render gig workers dependent on such unfair labour practices in the first place (Hickson, 2023: 10). To some extent, this is a general feature of capitalist labour markets (Cicerchia, 2022; Gourevitch, 2014; Mosar, 2023; Vrousalis, 2021b). The latter rely on ‘necessitous circumstances’ compelling individuals lacking resources for their personal survival to sell their labour-power to capitalist employers (Gourevitch, 2014: 81). Keynesian labour market reforms during the 20th century in the Global North have, however, alleviated the structural domination of some sectors of the capitalist labour markets by introducing social safety nets for workers. By providing workers with recognised union representation and secure exit options, welfare states have ameliorated workers’ bargaining position vis-à-vis employers. Without the direct threat of immediate deprivation, workers can negotiate better working conditions. Social protections like unemployment benefits and public healthcare have softened workers’ submission to the vagaries of the market. However, as Cole would highlight (Cole, 2020: 99 & 126), such measures only alleviate structural domination without eliminating it. The Keynesian interventions of his day granted workers minimal living standards through redistributive policies, but not a complete social transformation (Cole, 2017a: 182). Genuine freedom demands more than some antipower vis-à-vis labour market domination and employers. It requires genuine economic democracy and worker self-management.

However, gig workers often lack access to even these minimal social safeguards. Platform companies prefer to designate gig workers as self-employed to circumvent the protective measures that improve workers’ bargaining position. Such direct exposure to private labour markets enables platforms to pursue their agenda with little opposition. Structural domination from capitalist labour markets thus returns in full force in the gig economy. And since gig workers often lack the qualifications or opportunities for accessing the traditional labour market and its social protections, they are structurally impelled to linger in the gig economy with few exit options (Bieber, 2022: 9; Rogers, 2023: 107). National migration regimes regularly exacerbate this precarity (Schaupp, 2022: 320). Because many labour platforms forego stringent application procedures, migrant workers with precarious residential status often find their way into the gig economy. They start from situations of ‘compromised agency’ (Van Doorn, 2023: 173) in which labour platforms offer the best work realistically available, despite all its shortcomings, since the alternatives are worse jobs in non-digital areas of the gig economy or no income at all (Woodcock and Graham, 2020: 31). The same diagnosis applies to individuals with specific domestic care duties in their private lives, like single parents (Dubal, 2019: 18). For them, labour platforms and their scheduling flexibility might be the only viable option. Given the decline of health- and childcare institutions in some countries, these individuals might be unable to combine domestic care duties with traditional employment. The digital gig economy is then the only option left. Labour platforms thereby operate as de facto monopsonists on the labour market for particular workers (Nathan et al., 2023: 377; Van Doorn, 2023: 165–166). On the demand side of the labour market, there are only private platforms willing to hire these precarious workers. As Schor (2020: 49) notes, the more dependent individual workers are on labour platforms for their daily subsistence, the more problems they tend to experience interacting with them.

Labour platforms subsequently use multiple techniques of labour lock-in to consolidate worker dependence. For example, they restrict the portability of reputation data to ensure that workers cannot leave the platform without losing their data (Casilli, 2019: 323). Uber is also known for granting loan and leasing programmes for drivers to buy new cars (Ravenelle, 2019: 73; Rosenblat, 2018: 65). Uber drivers subsequently pay back these loans with exclusively app-generated revenues, which hinders them from leaving the platform without incurring tremendous debt. Uber also strategically underprices its services whenever it arrives in a new city to acquire monopoly status in the urban transportation sector (Rosenblat, 2018: 168–169). It subsidises rides with investor funds in order to attract drivers from traditional taxi companies and make competitors go out of business. When this endeavour succeeds, Uber is the only company left for taxi drivers, which gives it leverage to decrease wages. In New York City, for instance, Uber's base fare in 2014 was 6 dollars but went down to 2.55 dollars by April 2018 (Ravenelle, 2019: 75). Platform companies aim, in other words, to become simultaneously monopsonist vis-à-vis workers, in order to extract beneficial employment deals, and monopolist vis-à-vis consumers, in order to raise prices for its services and extract data from on-platform transactions. The platform becomes an inescapable intermediary for drivers and consumers to meet, with the platform company extracting rent from occupying this position (Couldry and Mejias, 2019: 129; Sadowski, 2020: 568). Both workers and consumers become dependent on the platform for their daily subsistence (Muldoon, 2022b: 13; Schneider, 2018: 321).

G.D.H. Cole and guild socialism

G.D.H. Cole constitutes a major source of inspiration for labour republicanism (Muldoon, 2022b: 81). In the early 20th century, British socialism was divided between Fabian and guild socialist currents. ³ The Fabians believed in a collectivist planned economy under state control. Guild socialists like G.D.H. Cole objected that this would merely shift relations of domination from capitalist management to state bureaucracies without establishing a genuinely participative industrial democracy (Cole, 2017b: 113). They favoured the Owenite experiments in cooperative governance, yet they distanced themselves from Owen's own paternalistic attitudes toward workers (Harcourt, 2023: 75–77). What makes Cole's guild socialism particularly helpful for the exploration of labour republicanism today is his insistence on domination in capitalism and the need for direct and collective worker self-government through a revival of the medieval guild system (Cole, 2017a: 44; Muldoon, 2022b: 81). Writing at a time when 19th-century laissez-faire liberalism was coming to an end, Cole applauded the rise of trade-unions struggling not just for collective bargaining rights but to take over factory management entirely. ‘When the industrial unrest took form, it was found to be not merely a demand for higher wages, but an insurgence against tyranny and an aspiration towards industrial self-government’ (Cole, 2017b: 104). Worker-managed cooperation did not just constitute antipower but a route to economic democracy.

Cole's focus on wage-slavery and republican freedom was not unique among socialists of the 19th century. In the United States, the Knights of Labor voiced a similar critique of capitalism linked to the promise of a ‘cooperative commonwealth’ similar to Cole's guild-system (Gourevitch, 2014; Scholz, 2023: 117–122), while pre-Marxist socialists in France, like Fourier and Proudhon, had expressed similar viewpoints (Harcourt, 2023, 77–82; Lamb, 2019: 65–68). What makes Cole interesting in the movement of socialist republicanism is, surprisingly, his anachronism. He defended labour republicanism when Leninist Marxism became the ruling ideology of socialist thought. Cole deemed any violent revolution in the West unrealistic and opposed a Leninist dictatorship of the proletariat (Cole, 2017a: 187–188; 2020: 171–172). He was critical of national centralised planning and Taylorist industrial technology as tools for proletarian progress. Cole pleaded for a decentralised economic system of worker-managed factories in which technologies would be used not to increase labour productivity but to enhance the quality of work. These features made Cole untimely in the early 20th century yet helpful for debates about the gig economy today. Protests against domineering labour platforms today rarely aim for global social revolution or central planning. They voice Colean aspirations for direct self-government, locally anchored worker autonomy, and technologies serving rather than undercutting meaningful work.

The philosophical background of Cole's stance is political associationalism (Cole, 2023). Influenced by Rousseau's social contract theory (Lamb, 2005), Cole believed that the individual will is the basic unit of social life, yet this will is always already ‘born into a complex society’ (Cole, 2023: 2) where it spontaneously forms associations with other individual wills. In contrast to Rousseau's notion of the general will, Cole never argues that these wills coalesce into one single collective will embodied by the state. Society rather remains a patchwork of overlapping communities with each possessing their own collective will, while individuals remain obligated to many different groups (Cole, 2017a: 12; 2020: 34). The best political framework for accommodating this diversity of associations is a functional rather than territorial democracy (Lamb, 2019: 93; Wright, 1978: 232). Individuals have different opinions about different subjects depending on their associational allegiances (Masquelier and Dawson, 2016: 6). Democracy should hence not try to represent the whole person at once based on where they live, but should foster mechanisms that make individuals participate in all their particular associations. Citizens attain freedom not through all-round representation in national parliaments but through institutional channels that stimulate collective self-government at work, in the home, in the neighbourhood, etc.

In the workplace, political associationalism entailed direct worker control over the factories and forming guilds in which workers would democratically coordinate production norms and prices (Cole, 2017a: 68). While guilds do not emerge ex nihilo worker-managed cooperatives and trade-unions would serve as foreshadowings of guild-based worker self-management (Cole, 2017a: 202). Guild socialism stood for a decentralised economic democracy where most governmental autonomy was located on the level of individual factories, where workers would most effectively exercise governmental power (Cole, 2017a: 47). Coordination between factories would gradually move upward to the level of national guilds, with representatives at each level elected from lower levels with an imperative mandate to functionally represent their constituencies on specific issues. While this counters relational domination, guild socialism also supplies a response to structural domination and economic precarity by offering income security to guild members (Cole, 2017a: 74). Like their medieval counterparts, modern guilds would pool resources together to offer a social safety net for members falling out of employment. Even when guild workers lose their jobs, they can still count on mutual solidarity to receive social protection. In this way, Cole's guild socialism not only confronts the relational domination of the capitalist workplace, but also the structural domination of economic insecurity plaguing precarious workers.

Inspired by the aesthetic socialism of William Morris, Cole believed that industrial self-government via guilds would revive the values of craftmanship connected to medieval guild culture (Cole, 2017a: 46; 2020: 90). In capitalism, workers do not reach their full creative potential due to relational and structural domination. Instead of fully determining their own conduct, workers are reduced to the role of subordinate executants of capitalist command. The Taylorist separation of conception from execution concentrates power and knowledge in the hands of management, while dispossessing workers of the capacity to establish expertise in their craft (Masquelier and Dawson, 2016: 8; Wright, 1978: 236). They submit to capitalist command to earn a living and avoid economic insecurity, but their labour does not express their own conceptions. Cole argues that industrial machines operate as conduits for capitalist domination by deskilling labour. The capacity for conception is concentrated in a managerial elite aiming not for qualitatively good craftsmanship but quantitatively cheap mass production. In worker self-government, Cole saw an opportunity to abolish the separation of conception from execution. Workers would control both the creative and executive dimension of the labour process. The modalities of the final product would be determined by workers’ own standards and desires. According to Cole, this would facilitate the full expression of workers’ creative capabilities, leading to a revival of artisanal production methods. Still, Cole was not technophobic (Cole, 2017a: 76; 2017b: 241). He even argued that technological automation could liberate workers from the toil of particular forms of dull or dirty work resistant to artisanal upgrading. While he argued against ‘the triumph of the man-machine’ (Cole, 2017b: 252), technology could be repurposed to enhance workers’ freedom to autonomously coordinate the labour process.

Later in life, Cole's views shifted more toward Fabianism in response to the failures of the guild-socialist movement. Existing guilds in the UK had failed (Cole, 2020: 117; 1944: 290–291) and Cole was deeply disappointed in the moderate liberal course of the first Labour government under Ramsay MacDonald (Cole, 2020: 98–99). While he never repudiated guild socialism, he downplayed decentralisation and the utopian elements of guild socialism in these later writings. In the name of political pragmatism, his politics moved toward the welfare-statism of Fabianism and the Labour party (Riddell, 1995: 938). New realities necessitated pragmatic shifts in politics. I will, however, focus on Cole's writings from the 1910s and early 1920s, when he most forcefully defended guild socialism. They provide the clearest account of how cooperatives can be mobilised to tackle domination under capitalism. Cole's strategic thought about the move from capitalism to guild socialism is, in this context, particularly inspiring for the platform cooperativist movement today.

Guild socialism goes online

The cooperativist movement often aligns with Cole's artisanal labour republicanism. Rather than reformist antipower, cooperatives constitute an island of worker self-management amid a capitalist milieu of relational and structural domination. The 19th-century Knights of Labor has already argued for a ‘cooperative commonwealth’ to gradually displace the capitalist ecosystem of private companies (Gourevitch, 2014: 123). Cooperatives give form to working-class solidarity as a new basis for social organisation, and their strategy is to scale up this counter-hegemonic ecosystem of working-class institutions until they supplant the status quo (Gourevitch, 2014: 164). Cole rejected the usual socialist dichotomy between reform or revolution in favour of this exodus strategy (Cole, 2017a: 174): while social reforms would merely ameliorate workers’ bargaining position without supplanting capitalist domination itself, revolutionary action was unrealistic for the Western working class. ‘The thing to aim at […] is not early revolution, but the consolidation of all forces on the lines of evolutionary development with a view to making the “revolution” […] as little as possible a civil war’ (Cole, 2017a: 187). Cole wished to strengthen working-class capacities for direct self-management so thoroughly via the guild-system that the latter could gradually displace and supplant capitalism, a transition he dubbed ‘a policy of encroachment’ (2017a: 196). While he still believed revolution to be necessary, the latter would have a relatively minor role in the grander scheme of building an alternative society within and against capitalism.

Contemporary platform cooperativism has expressed similar sentiments with calls for ‘intercooperation’ (Grohmann, 2023: 4), ‘cooperative scaling’ (Scholz, 2023: 50), and the ‘building of radical democratic institutions within the “cracks” of the capitalist system with a view on eroding the dominance of capitalist institutions’ (Muldoon, 2022b: 148). The Indian union-cooperative Self-Employed Women's Association (SEWA), for example, started as a union defending poor self-employed women in India's informal economy (Scholz, 2023: 127–130). When the organisation noticed that members experienced difficulties accessing loans, they established a cooperative bank, and when similar issues emerged for health insurance, they created a mutual fund. With this scaling strategy, SEWA has grown a network of cooperatives that strengthen members’ capabilities for self-determination across a wide domain of social activities. Such proposals share a strategy for circumventing the ‘reform-or-revolution’ dichotomy by investing in institutional ecosystems that operate within and against capitalist environments of domination. While my argument focuses on worker-managed platform cooperatives as a hotspot for such a cooperative commonwealth, its ultimate object is hence economic democracy in general.

Labour platforms are an interesting starting point because advances in platform communication have made it easier for people to spontaneously organise online without the need of explicit hierarchies (Benkler, 2006; Castells, 2015; Scholz, 2014). Amazon Mechanical Turk workers have, for instance, organised political campaigns via Reddit forums and specialised communication platforms (Della Porta et al., 2023: 109–110), while Deliveroo couriers organised their first UK strikes with Whatsapp-groups and social media campaigns (Cant, 2020: 130–133; Ovetz, 2023: 286). Cooperative platforms are part of this movement toward worker self-organisation. Cooperatives recode platform technology to offer alternatives to the privately-owned platform companies of the gig economy. For every famous labour platform, there are cooperative alternatives currently available. Deliveroo nowadays competes with CoopCycle, Uber with Juno and Drivers Cooperative, Airbnb with Fairbnb. Platform technology can be reappropriated and recoded to serve worker self-government rather than dominating workers through Taylorist methods of labour coordination (Casilli, 2019, 379; Muldoon and Raekstad, 2023, 13; Zygmuntowski, 2018: 181). Since worker cooperatives jobs are persistently more stable and offer better working conditions than those at other companies (Pérotin, 2013; Scholz, 2014), cooperativism not only addresses the relational domination of digital Taylorism, but also the structural domination from economic precarity. Workers at platform cooperatives are more assured that their job is relatively secure and they will not be arbitrarily deactivated.

But this shift requires a different approach to platform design, echoing Cole's critique of Taylorism. As mentioned, software developers often embed and hide the separation of conception from execution in seemingly neutral app design (Pasquale, 2020: 23; Rogers, 2023: 74). Through unilateral design choices, platform technologies predominantly benefit corporate revenue maximisation, rather than workers’ well-being and autonomy. Technology design must involve workers themselves in platform development if cooperativist data experts wish to avoid domination, even in its benevolent or paternalist guises (Scholz, 2023: 193). By creating strategies of co-creation in software development, cooperativist data experts can embed workers’ own aspirations into these technologies (Bauwens et al., 2019: 39; Salvagni et al., 2022: 712; Scholz, 2017: 181). CoopCycle, for instance, is a food delivery initiative that connects local cooperatives together in a federative network. CoopCycle itself only provides the coding for building an app and its expertise to tailor this platform to the specific needs of local worker cooperatives. The eventual platform is hence the product of collaboration between centrally dispatched experts and the local groups themselves. CoopCycle allows worker cooperatives to adapt the algorithms to their own specific context but offers help from in-house IT experts on the technicalities. In Cole's terminology, CoopCycle functions as a food delivery guild that coordinates the work of a multitude of local cooperatives with each maximum local autonomy.

Like in Cole's argument, platform cooperativism's worker self-government facilitates the full self-expression of workers’ creative potential, which explains the frequent comparisons between platform cooperatives and medieval guilds (Bauwens et al., 2019: 57; Scholz, 2017: 169). While workers in the gig economy often showcase ‘embryonic solidarity’ (Tassinari and Maccarrone, 2020: 39), cooperative platforms create opportunities to articulate this affective solidarity into a shared identity and community of workers sharing a craft (Schor, 2020: 149). Stocksy United, a platform cooperative for stock photo photographers provides a good example (Schor, 2020: 164–168; Schor and Eddy, 2022). It was founded by former owners of iStockphoto, a small company sold to Getty Images in 2006. But when the overall quality of the photographic work dropped dramatically, they regretted the deal. These frustrations with the lack of craftmanship in favour of mass production formed the basis for a cooperative alternative where photographers would set their own quality standards. Stocksy's members are located across the world, so they communicate with each other and shape Stocksy's policies via online discussion forums. To evaluate the quality of photographic work, Stocksy relies mainly on the knowledge and shared values of the photographers themselves. At first, for instance, Stocksy set an upper limit of 1000 members, so potential applicants had to prove themselves before getting a place at Stocksy. The joke even circulated that it was easier to get into Harvard than Stocksy United (Ravenelle, 2022). The company selects only the best portfolios to distribute on the platform. That decision depends not on who can take the most photos for the least amount of money, but who can deliver the best quality. After all, as experts in the business, photographers are particularly good judges of quality.

Objections to platform cooperativism

Platform cooperativism is not without its critics and, admittedly, even when they succeed they rarely grow as large as their private competitors. In this final section, I would like to focus on three common objections. The parasitism objection claims that worker cooperatives tend to maximise their own interests at the cost of other stakeholders, mainly consumers. Platform cooperativism would hence reduce domination for gig workers, but also create new relations of domination between worker cooperatives and their clients. After dealing with this common criticism I turn to the chief criticism against cooperativism, the so-called ‘degeneration thesis’ (Cornforth, 1995: 487; Unterrainer et al., 2022: 2). Originally formulated by the British Fabians Beatrice and Sidney Webb, it states that democratically organised worker cooperatives tend to degenerate into hierarchically managed capitalist firms over time (Webb and Webb, 1914). This argument has two dimensions (Diamantopoulos, 2012: 201). The Luxemburg objection focuses on the external pressures from capitalist market competition to adapt to the exploitative practices of capitalist forms to survive, while the Michels objection stresses pressures internal to cooperatives for new, implicit hierarchies to emerge.

The parasitism objection

Some object to worker-owned cooperatives because these would be parasitic on other stakeholders, mainly consumers (Bieber, 2022: 9). Medieval guilds infamously defended their revenues at the cost of consumer welfare with anti-competitive protectionism. One could expect modern guilds to behave similarly. By putting workers in control of the production process and prices, cooperatives can overcharge consumers for suboptimal services (Cole, 2017b: 281). Especially in markets already tending towards monopolisation, like the platform sector, this is a serious risk. However, Cole (1944: 285) explicitly presents guild socialism as a reconciliation of syndicalist worker self-management and consumer concerns. Yet his arguments are often unconvincing. His first counter argument (2017b: 256) is that workers would refrain from duping consumers, because they themselves are also consumers, so they would be exploiting themselves. While this could be true for some basic social goods, like food production or energy provision, this response does not work for the majority of economic goods. Someone who works in car manufacturing does not necessarily own a car, so this individual could still easily exploit consumers at no personal cost.

Cole's second counter argument relies on a naively optimistic account of human nature (Wright, 1978: 229). Cole argues that the inclinations toward free communal service and meticulous craftmanship are inherent to human nature itself (Cole, 2011: 1d; 2017b: 256). While capitalism thwarts these innate impulses to serve the profit motive, guild socialism allegedly unleashes the human creative drive without distortion. The cooperative form hence aligns with human nature and the latter supposedly guarantees that workers can be trusted ‘on their honour’ (Cole, 2017a: 88). ‘Human nature’ operates in this counter-argument as a deus ex machina to solve the parasitism objection, but there is little evidence to put absolute trust on workers’ innate altruism. While the philosophical tradition of cooperativism sometimes relies on evolutionary arguments for cooperative inclinations, biological evidence is not as conclusive (Harcourt, 2023: 90). However, endorsing the opposite assumption, that is, innate egoism, would also be misleading. The problem with Cole's argument is not the reference to altruism per se – human beings are indeed able to act altruistically – but its attempt to ground human inclinations in a nebulously defined yet supposedly eternally stable ‘human nature’. If the optimistic hypothesis of innate altruism is unconvincing, then so is its pessimistic opposite of irredeemably egoistic homines oeconomici. Worker cooperatives are naturally inclined neither toward free communal service nor to consumer exploitation.

This observation leads to Cole's third and most convincing counter argument. Rather than locating the inclination toward communal solidarity in human nature, Cole sometimes attributes it to ‘the social mechanism’, that is, social institutions (Cole, 2017a: 25–26). Workers’ institutional environment can alter their conduct and thereby govern them in the direction of both egoism or communal service. The point of guild socialism is to ‘breed free men [sic]’ (Cole, 2017a: 61) rather than trusting human nature to spontaneously run its course. Companies that motivate workers through a combination of coercion and pay breed subservient individuals, while institutions that call upon workers’ solidarity to the community and their sense of pride in a good day's work can expect workers to respect communal and artisanal responsibilities. Cole's institutional answer to the parasitism objection (2017a: 88) is thus to establish coordinative meetings between producer and consumer representatives. Their function would be to host a deliberative forum for the setting of prices and production norms. In more contemporary parlance, Cole's answer constitutes a call for deliberative stakeholder democracy to avoid power imbalances between worker cooperatives and their consumers. ⁴ Producer and consumer representatives should form deliberative councils in which they determine prices and product standards by consensus (Cole 2017a: 89).

In practice, Cole (2017a: 80) opposes the formation of one big consumer association since this would forego the diversity of political associationalism and the differences among consumer goods and profiles. In Guild Socialism Restated, Cole distinguishes between standardised collective goods, for which municipally anchored Collective Utility Councils can represent consumer interests, and more specific individual consumer goods, which are best represented by multiple consumer cooperatives (Persky and Madden, 2019: 441). Mirroring the decentralised organisation of the guilds, these consumer organisations would both inform guilds of consumer preferences and protect consumers from monopolistic abuses. In Self-Government in Industry (2017b: 283), Cole adds a tax proposal that would automatically tax away potential monopoly rents from guilds, though the proposal remains vague. For today's labour platforms, Cole's proposal of deliberative forums for producers and consumers is helpful, yet not necessarily in the exact same format. Just as Cole often pleads for pragmatism depending on which institutional associations offer effective member participation, so too platform cooperatives should remain open-minded about how to include consumers in their internal decision-making. In later texts, Cole criticises some consumer cooperatives that failed to institutionalise active participation of their members (Cole, 1944: 386). We should not make the same mistake by pre-determining deliberative channels for consumer interests in advance. Fairbnb, for example, offers consumers a voice in platform management via municipal governments that articulate the concerns of tourists and local residents. On the other hand, Signalise, a platform offering interpreting services to deaf people, integrates consumer concerns by making deaf clients co-members of the cooperative itself. Both strategies have their merits, and the choice would depend on which offers the highest chance for equal and fair deliberation among producer and consumer representatives.

The Luxemburg objection

Marxist critics of worker cooperatives focus on the market pressures coming from capitalist competitors (Papadimitropoulos, 2021; Sandoval, 2019). Though Marx acknowledged the contribution of cooperatives in prefiguring post-capitalist futures (Ranis, 2016: 2), he believed worker cooperatives would ultimately be ‘forced to become their own capitalists’ (Marx, 1992: 571). As Rosa Luxemburg elaborated, workers ‘are obliged to play the role of capitalist entrepreneur toward themselves—a contradiction that accounts for the usual failure of cooperatives in production, which either become pure capitalist enterprises or, if the workers’ interests continue to predominate, end by dissolving’ (Luxemburg, 1999). While cooperatives address the relational domination from the employer by championing worker self-government within the firm, they ignore the ‘mute compulsion of the market’ on these firms (Mau, 2023). In capitalist free markets, companies can only survive by competing with other companies. If those competitors marshal relational domination toward workers to maximise their revenues, cooperatives either have to follow suit or be left behind (Du Toit, 2019: 8; Woodcock and Graham, 2020: 139). Later in life, when the UK builders’ guild collapsed, the disappointed Cole even somewhat agreed with this criticism (Cole, 1944: 286). More than in his previous work, he then stressed the need for state-led nationalisation of companies as a first step toward guild socialism (Cole, 1944: 292; 2020: 94). ⁵ While the economy should still be managed through a decentralised guild system, Cole only envisages this system's success if the state first appropriates the means of production by force to avoid the guilds’ erosion by capitalist competitive pressures. Platform cooperatives generally do not take this position.

However, in the case of labour platforms, there is little chance of keeping a cooperative island afloat amidst a sea of monopoly-aspiring companies. It will either be overrun or it will have to mimic the practices of its domineering competitors. However, the platform sector is not a traditionally competitive free market. Private platform companies use investor funds to outcompete others through strategic underpricing and exploiting legal loopholes. Uber, for instance, famously evades normal taxi regulations in several countries to operate at lower costs, claiming it is not a taxi but a technology company (Rosenblat, 2018: 5). The two implications of this practice are that (a) not even (uncontroversially) competitive companies can endure against private platform companies that can afford to run at financial losses for years, and (b) private platform companies can only hope to remain competitive in the long run if they influence the law and acquire a monopoly before the investor funds run out. In that situation, it would be unwise to engage in direct synchronic competition.

Rather than trying to overtake private platform companies immediately, cooperatives can better convince governments of the need for stricter sector regulations, like employee status for gig workers, while waiting for the latter to run out of money. Diachronic competition is the preferable strategy. Cooperative platforms can step in once a private platform fails to secure market dominance. The case of BestellenBij from the introduction illustrates this strategy. The worker cooperative filled a void once Deliveroo had left the Dutch market. Rather than directly trying to outpace Deliveroo, BestellenBij waited for Deliveroo itself to fail due to the reclassification of its couriers. An even clearer example comes from the city of Austin in Texas, where Uber and Lyft decided to leave the urban transportation market over a regulatory dispute with the municipal government (Muldoon, 2022b: 113–114; Solel, 2019: 240). The companies had hoped for politicians to change their minds once drivers and consumers started complaining about Uber and Lyft's departure. However, the municipal government had anticipated this scenario and invested in a non-profit called ‘RideAustin’ and a worker cooperative called ‘ATX’. Workers and consumers quickly shifted to the new platforms. This example shows that platform cooperativism is a sustainable organisational form, granted that these cooperatives constitute diachronic successors to, rather than synchronic competitors of, privately-owned platforms. Uber and Lyft were only able to return to Austin after lobbying with the Texas government to overrule the city's legislation.

The Austin example also shows that platform cooperativism requires a financial and regulatory framework supportive of worker cooperatives to ensure their emergence and long-term survival (Bunders et al., 2022: 3; Ranis, 2016: 36; Rojas, 2017: 195), as Cole also stipulated (2017a: 67; 1944: 287). In his discussion of worker cooperatives in Guild Socialism Restated, Cole explicitly acknowledges that ‘it is simply not possible […] to drive the possessing classes out of industry simply by competing with them under conditions which these classes themselves prescribe’ (2017a: 191). If the Austin government's regulation and investment choices had not first undercut the big platforms’ business model, the ousting of Uber and Lyft would have been unimaginable. The difficulty of accessing start-up capital is one of the central hurdles for cooperative platforms today (Bunders and De Moor, 2023; Rojas, 2017). Some cooperatives can survive through crowdfunding (Zygmuntowski, 2018: 184) or by servicing niche markets neglected by the big platforms (Salvagni et al., 2022: 711), but most need state support (Bria, 2017: 220; Grohmann, 2023: 7; Schneider, 2018: 333). Municipal governments, as in Barcelona, often play a crucial role by setting up investment funds and support networks for local cooperatives (Muldoon, 2022b: 101; Salvagni et al., 2022: 715; Schor, 2020: 172). They create not only a regulatory environment in which platform cooperatives can withstand competition from privately-owned platforms, but they also provide start-up capital to finance a cooperativist ecosystem. Such initiatives nurture a democratic voice in the investment decisions that determine the course of technological innovation. In terms of the republican agenda of freedom as non-domination, creating such a public investment strategy only supports the struggle against the structural domination of precarious labour markets (Mazzucato, 2018). While private investors often search for technological innovations that promise to cut labour costs, public investment funds can prioritise other, more democratic aims (Scholz, 2017: 186).

The Michels objection

The final objection to platform cooperativism is that internal pressures erode the democratic aspirations of cooperatives (Diamantopoulos, 2012: 204). Power purportedly gets concentrated in new implicit elites (Schneider, 2022). Cooperatives often start as democratic experiments but regularly devolve into hierarchical institutions because some members hoard decision-making power while others refrain from actively participating in everyday governance practices. In his 1911 classic Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy, Robert Michels termed this ‘the iron law of oligarchy’ (Michels, 1968: 342). This criticism is, for example, regularly levelled at Mondragon, one of the largest and most famous worker cooperatives in the world (Bretos et al., 2020; Kasmir, 1996). This company started as a democratic manufacturing cooperative in the 1940s, but grew so large over time that new hierarchies between management and workers eventually emerged. The separation of conception and control that worker self-government was supposed to overcome returned to deal with the complexity of a large business. Especially for platform cooperatives – where there is from the start a divide between digitally highly-skilled experts and workers – the risk of such a separation emerging is considerable (Bunders, 2021: 201). The separation between conception and execution can quickly return when the cooperative platform remains a black box for workers, while data experts determine unilaterally how the platform software coordinates the labour process.

However, empirical research on cooperative governance shows that the law of oligarchy is made of a metal more pliable than iron. Unterrainer et al. (2022: 9) have shown in their systematic review of cooperative governance that less than 10% of cooperatives degenerate in the way critics suggest. The majority either retains the original democratic framework or regenerates after a degenerative phase. A more realistic approach describes cooperatives as oscillating between oligarchic and democratic tendencies (Cornforth, 1995: 487; Kasparian, 2022: 210; Unterrainer et al., 2022: 22), a view Cole (1944: 389) approximated in The Co-Operative Movement but never fully developed. Different cooperatives occupy divergent positions on a spectrum between oligarchical and democratic extremes, and they continuously shift positions over time. While it takes energy and hard work to retain power among the workers, it also requires effort from managerial elites to amass the power to dominate others within a cooperative. These two forces are locked in a continuous struggle over the soul of the cooperative. Institutional measures, like job rotation, can help in safeguarding the internal democracy of worker cooperatives (Cornforth, 1995: 499). By ensuring that members can only occupy positions of power for well-delineated tenures, while everyone regularly rotates into such positions, members are encouraged to reflect on company decisions simultaneously from the perspective of workers and managers. Channels for official contestation from workers vis-à-vis managerial decisions also support a democratic culture in the cooperative (Chaves and Sajardo-Moreno, 2004: 140; Kasparian, 2022: 114; Unterrainer et al., 2022: 24). Worker cooperatives thereby become antagonistic spaces where different power relations are played out without ossifying into a Michelsian oligarchy (Christiaens 2023a: 159; Kasparian, 2022: 216).

Another requirement already stressed by Cole himself is the need for a ‘spiritual revolution’ among cooperative members (2017b: 277). If cooperatives wish to remain sites of worker self-government, they have to promote subjective and affective commitment among workers to the shared values of the community of the cooperative (Cole, 2017b: 103). This entails not just passive acceptance from workers to managerial decisions, but civic virtue and active citizenship (Cole, 2017a: 12). The literature on platform cooperativism and cooperativism in general stresses the need for a common identity, shared values and subjective commitment (Bauwens et al., 2019: 39; Gourevitch, 2014: 153; Rothschild-Whitt, 1979: 520; Unterrainer et al., 2022: 16). Democracy takes time and is emotionally demanding (Chaves and Sajardo-Moreno, 2004: 144; Harcourt, 2023: 46; Mannan and Pek, 2023: 16–17). It is hence crucial that members are ready to commit to the cooperative and its principles if the tendency toward oligarchisation is to be halted. In New Zealand, for instance, the cooperative network Enspiral invests heavily in member education to ensure the continuance of its philosophy among members. Enspiral started in the wake of Occupy Wall Street New Zealand, but today it connects several cooperatives, like a freelance accounting platform (Enspiral Accounting), open-access, decision-making software development (Loomio), and a web-development school (Dev Academy). All members and partner organisations have to share the ethos and commitments of the original OWS inspiration to enter the community. Enspiral communication consistently talks of ‘high-trust relationships’, ‘shared values’, and ‘the collective personal sense of commitment’ (Pazaitis et al., 2017). Since most people mainly devote time and energy to initiatives to which they are emotionally committed, it is crucial for cooperatives with diverse crowds to encourage these forms of emotional involvement and even demand it as a requirement for hiring.

Conclusion

Initially, we wondered how the withdrawal of Deliveroo and Helpling in the Netherlands could have been resolved more elegantly. The reclassification of gig workers had granted them antipower against the domination of labour platforms and economic precarity. However, especially with Helpling leaving the market entirely, domestic workers were worse off than before. In this paper, I have argued that the central obstacle to gig workers’ freedom is domination. From a relational perspective, gig workers are subjected to labour platforms that impose algorithmic techniques of digital Taylorism to increase platform revenue at workers’ expense. From a structural perspective, precarious workers have few options left but to work for such labour platforms, since the latter act as monopsonists on the labour market. Relying on G.D.H. Cole's account of guild socialism, I have argued that platform cooperativism champions worker self-government and increased economic security for cooperative members. I have also defended platform cooperativism against three common objections. Some critics, firstly, stress the risk of parasitism vis-à-vis other stakeholders, like consumers, but such risks can be avoided with the right institutional framework to encourage deliberative stakeholder democracy. Marxists, secondly, warn of the external competitive pressures on worker cooperatives in capitalist free markets. However, examples like BestellenBij and the Austin case show that worker cooperative can diachronically compete with private labour platforms by supporting strict regulation of the gig economy and waiting for platform companies’ investment funds to run dry. The Michels objection, lastly, fears the return of internal hierarchies in worker cooperatives due to the digital divide between the data experts in control of platform design and the workers, who are often laypeople when it comes to technology development. Inclusive co-creative design, institutional measures and the cultivation of a cooperative mindset among workers can, however, halt the slide toward oligarchisation. Worker cooperatives should rather be understood as antagonistic spaces where tendencies toward the concentration and decentralisation of power are continuously played out.