2023 Trajectories of Platform Capitalism and Platform Work PhD Summer School 2022

PHD SUMMER SCHOOL 2022 TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK

Contents FOREWORD 2 1 WORKERS’ EXPERIENCES OF UNPAID PLATFORM LABOUR AS INTENTIONAL ACTION 3 Abstract ............................................................................................. 3 1.1 Introduction ........................................................................................ 3 1.2 Background ........................................................................................ 3 1.3 Theory and methodology ...................................................................... 4 1.4 Results ............................................................................................... 5 1.5 Conclusion ......................................................................................... 7 1.6 Acknowledgements ............................................................................. 7 2 COALITIONAL POWER IN THE DIGITAL ECONOMY: ALLIANCES OF GIG AND TECH WORKERS 9 Abstract ............................................................................................. 9 2.1 Introduction ........................................................................................ 9 2.2 Background: Labour Struggles in Gig and Tech ......................................... 9 2.3 Methodological Approach ..................................................................... 10 2.4 Results ............................................................................................... 11 2.5 Conclusion ......................................................................................... 13 2.6 Acknowledgements ............................................................................. 13 3 PLATFORM WORK IN MADRID, MILAN, AND SAN FRANCISCO: EXPLAINING MUNICIPAL RESPONSES THROUGH A MULTI-LEVEL GOVERNANCE FRAMEWORK 16 Abstract ............................................................................................. 16 3.1 Introduction ........................................................................................ 16 3.2 Background ........................................................................................ 16 3.3 Theory and Methodology ...................................................................... 17 3.4 Results ............................................................................................... 18 3.5 Conclusion ......................................................................................... 20 1

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK FOREWORD The rollout of the platform economy has renewed the debate over the future of employment in the digital era by raising concerns about the working conditions of platform workers and the risk of precariousness it entails. While work on platforms resembles many work arrangements that have been around for some time, new issues have arisen in connection with platform work, especially in relation to(1) the unclear employment status of platform workers, with implications for their rights and obligations in terms of job and social protection, and(2) algorithmic management techniques used to assign tasks, evaluate worker performance, and compensation of workers. This current publication is based on three student papers as part of the PhD Summer School 2022 – Trajectories of platform capitalism and platform work. The summer school was organised by the FES Competence Centre on the Future of Work in Berlin, in cooperation with the Weizenbaum Institute for the Networked Society. By situating individual research projects within the broader research agenda on the platform economy and platform work, the summer school focused on a range of topics, including studies exploring algorithmic management on labour platforms; issues of race, gender and migration in platform work; varieties of platform capitalism; the role of platforms in reorganising traditional industries; ethics of digitalisation, and strategic foresights. The summer school brought together 14 PhD students from different European universities, e.g. from Germany, Italy, the Netherlands, Norway, Slovenia, Sweden, Poland and the UK, and was supported by five academic researchers: Prof. Dr. Martin Krzywdzinski, Dr. Florian Butollo, Dr. Tatiana Lopez-Ayala, Dr. Funda Ustek Spilda and Dr. Charalambos Tsekeris. The current publication looks at different levels of contention in platform work domain-based qualitative data analyses: unpaid labour across online freelancing, food delivery and domestic work platforms; collective action efforts of gig and tech workers and the role of municipal public institutions in mitigating platform workers’ working conditions. 2

Workers’ experiences of unpaid platform labour as intentional action 1 WORKERS’ EXPERIENCES OF UNPAID PLATFORM LABOUR AS INTENTIONAL ACTION by Milena Franke(KU Leuven, Centre for Sociological Research) and Valeria Pulignano(KU Leuven, Centre for Sociological Research) ABSTRACT This paper examines platform workers’ experience with unpaid labour by examining the constraints they face and the ways in which they intentionally use unpaid labour to navigate these constraints. Based on a qualitative study exploring online freelancing, food delivery and domestic work platforms in Belgium, we shed light on the intentional actions by workers, explaining how they use different resources to engage in unpaid labour and how this helps them to navigate constraints such as non-transparent work assignments and rating systems. We also show how workers face financial, private, physical, mental and career-related limits to their engagement in unpaid labour which can cause them to reduce their engagement or to exit the platform. Our aim is to contribute to an understanding of how the prevalence of unpaid labour on platforms fundamentally relates to workers’ actions and the resources at their disposal. Based on a qualitative study of three platforms upon which freelancing, food delivery and domestic work are performed in Belgium, we seek to deepen understanding of unpaid platform labour by focusing on workers’ experience with intentional unpaid activities undertaken by them to mitigate constraints such as non-transparent algorithms and rating systems. We furthermore identify a tension between worker engagement in unpaid labour involving commitment of their financial, private, physical, mental and professional resources, and the limits workers encounter because they are either unable(e.g. due to limited financial means) or unwilling(e.g. prioritising their private life) to commit. We argue that examining how workers intentionally use and/or limit the use of their resources helps to understand the conditions underlying their engagement in unpaid platform labour. 1.2 BACKGROUND 1.1 INTRODUCTION This paper explores why workers engage in unpaid labour and how they experience this engagement within different digital labour platforms, while focusing on the way in which they intentionally use unpaid labour to navigate the constraints imposed by platforms. Recent studies on platform work – that is, paid work mediated via on‐ and offline labour platforms – have revealed that platform workers perform unpaid labour, such as waiting, applying or preparing for work, on a significant scale(Pulignano& Marà, 2021; Shevchuk et al., 2019). Scholars have explained the prevalence of unpaid labour by highlighting the aspect of algorithmic control and precarious working conditions imposed by platforms, while unpaid labour is then made a requirement to access paid work(Pulignano et al., 2022b; Wood et al., 2019). Platforms cut costs by excluding‘unproductive’ labour from payment by means of piece-rate compensation(Moore& Newsome, 2018) and shifting demand-related risks to workers(Barratt et al., 2020). Therefore, there seems to be a clear motivation for platforms to push for unpaid labour. What is less well understood, however, is why and how workers engage as well as their understanding of unpaid labour. UNDERSTANDING UNPAID PLATFORM LABOUR THROUGH WORKERS’ EXPERIENCE AND WORKERS’ RESOURCES For the purpose of this paper, unpaid labour is defined as a productive but unremunerated activity that contributes to accessing and/or completing tasks within labour platforms (Pulignano et al., 2022b). Unpaid labour can take on various forms, including time-related(e.g. waiting time, communications with clients) and non-time-related forms(e.g. purchase of tools, payment of platform fees) and is likely to differ between platforms, depending on the sector in which they operate and the payment and rating systems employed (Pulignano& Marà, 2021). Recent literature points to unpaid labour as a means used by on-demand platforms to shift market uncertainty and costs to workers without offering them the protection provided by traditional forms of employment(Huws et al., 2018; Pulignano et al., 2022a). Platforms leverage unpaid labour with the aim of exploitation and commodification(Fuchs& Sevignani, 2013; Wood et al., 2019) and are able to coerce workers to undertake unpaid tasks through the use of algorithms and information asymmetries(Shapiro, 2020; Rosenblat& Stark, 2016). Workers actively manipulate platform rules, turning them to their own 3

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK advantage(Pulignano& Franke, 2022), however, and try to comprehend their experience of precarious conditions while at the same time being classified as‘free’ entrepreneurs (Josserand& Kaine, 2019). Precisely because of their unpaid nature, unpaid activities are difficult for platforms to control, so they tend to rely on workers’ initiative(Briziarelli, 2014). For example, workers’ unpaid search for jobs can extend across platforms, thereby potentially shifting their productive activities away from one platform(Gerber& Krzywdzinski, 2019). We therefore explore how and why workers actively engage in unpaid labour. Unpaid labour involves workers committing their own resources, such as their private time, material belongings or financial assets in the form of their household, or through employment elsewhere(Pulignano& Morgan, 2022; Holtum et al., 2021). Hence, workers might face constraints on their engagement in unpaid labour. Costs involved in accessing or carrying out platform work may prove to be unaffordable to some workers, causing them to exit or circumvent the platform(Maffie, 2022). The requirements of constant availability and the need to adapt their schedules to clients’ needs(Berg et al., 2018; Shevchuk et al., 2021) might be incompatible with workers’ other jobs or personal lives. Moreover, continuous exposure to job insecurity and task fragmentation can exhaust workers’ mental health(Glavin& Schieman, 2022) and may hinder their professional and skillset development (Shibata, 2020; Webster, 2016). It is with all this in mind that the present study considers both the resources deployed and the‘limits’ above and beyond which workers are unable or unwilling to engage in unpaid labour. THE PLATFORM ECONOMY IN BELGIUM The research was conducted in Belgium, a country with a small but growing platform labour force in a wide range of sectors, including both online freelance work, such as translation and IT services(Kässi& Lehdonvirta, 2018) and location-based work, such as food delivery or domestic services(FOD Financiën, 2022). The rise of the Belgian platform economy has been accompanied by the introduction of socalled‘peer-to-peer status’, exempting workers on registered platforms from tax and social contributions for earnings up to€ 6,340/year between 2018–2020. Since 2021, a 10.7% tax rate has applied because the Belgian Constitutional Court overturned the tax-free scheme(Paelinck, 2020). Platforms in Belgium also retain the possibility to hire workers as self-employed contractors or under‘student self-employed’ status, allowing students aged between 18 and 25 to earn up to€ 7,000/year from self-employed work without paying tax or social contributions. For the majority of workers, platform work is occasional and supplements other earnings, although a small percentage of workers rely on platform work as their main source of income(Drahokoupil& Piasna, 2019). on workers’ lives by extending value-extraction beyond the boundaries of paid work and capitalising on workers’ resources and the income provided by social protections and other employers(Srnicek, 2016; Walker et al., 2021). Yet, workers themselves manage constraints and develop practices to sustain their livelihoods in the platform economy (Anwar& Graham, 2020). A qualitative case-study research design of three labour platforms in Belgium was chosen to generate insights into unpaid labour within and across different platforms. Aiming to create variation regarding the experiences and forms of unpaid labour, we selected platforms employing piece-rate payment systems with different occupations(Pulignano& Marà, 2021) and that are as diverse as possible with regard to spatial dispersion, skill level and complexity of work(Vallas, 2019). First, we selected the food delivery platform Deliveroo. Deliveroo hires couriers under the(student) self-employed and peer-to-peer status and pays self-employed (student) couriers a distance-based, and peer-to-peer couriers, a fixed fee 1 . Couriers can log in to their apps whenever they want and are assigned orders that have been received through an algorithm based on their location and expected preparation times at restaurants. Second, the domestic work platform Ring Twice was selected, which provides diverse services such as babysitting, gardening, house repairs and many others. Ring Twice workers are peer-to-peer or self-employed and access work by applying to job posts on the platform. Workers stipulate an hourly or fixed-price pay rate when they apply, and the platform releases payment upon completion of the job and confirmation by the client. Ring Twice ranks workers according to‘experience levels’ – based on client ratings, identity verification, number of return customers, jobs completed and seniority on the platform – and calculates a ‘realisation percentage’ measuring the proportion of realised relative to cancelled jobs. Third, Upwork is an online freelancing platform in which we selected IT, graphical design, translation and copywriting freelancers. Belgian Upworkers are self-employed(students) or do not declare their platform activity. When posting a job, clients specify how much they are willing to pay and freelancers can bid below, but not above, the stipulated price. Upwork uses a rating system consisting of client reviews and a job success score that monitors the rate of successfully completed platform jobs. We conducted 30 narrative interviews with platform workers and three semi-structured interviews with platform managers between spring 2020 and autumn 2021 and complemented these interviews with desk research(platform websites in particular). We used a purposeful sampling strategy, diversifying respondents with regard to gender, age, employment status and the extent to which they relied on other sources of income. The data was analysed applying the Gioia method (Gioia et al., 2012). 1.3 THEORY AND METHODOLOGY In answering questions about why platform workers engage in, and how they view unpaid labour, we draw on conceptualisations of unpaid labour as both‘captive’ and potentially ‘free’(Terranova, 2000). Platforms put a significant strain 1 Deliveroo applies differently structured fees for peer-to-peer couriers because the Belgian tax authorities have challenged platforms’ use of peer-to-peer contracts. By fixing fees and showing that clients can know up front how much they will pay for the delivery, Deliveroo was able to justify the continued use of peer-to-peer status. 4

Workers’ experiences of unpaid platform labour as intentional action 1.4 RESULTS UNPAID LABOUR AS A RESULT OF CONSTRAINTS IMPOSED BY PLATFORMS Workers on all three platforms feel obliged to undertake unpaid labour, as platforms confront them with significant challenges. First, platforms constrain workers’ access to work and/or clients. By increasing the size of the workforce and employing piece-rate payment systems, platforms obtain unpaid labour through competition for limited tasks:“you get paid per delivered order, so if you only get one order per hour, you earn very little”(BEMF35). On Ring Twice and Upwork, rating systems further limit workers’ access to work, as “clients usually select someone with a lot of stars”(BEMF22). Deliveroo couriers report that“the order assignment system is non-transparent”, resulting in unpaid waiting time between orders because“you never know who will receive an order”(BEMF38). All three platforms constrain workers’ access to pay in a similar way. Ring Twice and Upwork charge a commission on all platform income, whereas Deliveroo unilaterally lowers and restructures fees, as was experienced by peer-to-peer couriers whose fees were fixed, requiring them to devote unpaid time to delivering long-distance orders. Freelancers and domestic workers report that they“have to lower(their) prices” due to strong price-based competition(BEMF24): “I started with my rates at$ 30 an hour and then I realised:‘This isn’t working out, that’s too expensive.’ I had to cut them to$ 20”(BECM01). Moreover, food delivery and domestic workers incur costs to purchase equipment such as bikes and tools. As peer-to-peer workers are excluded from social security, they are unable to access sick or holiday pay. A peer-to-peer courier who recently underwent surgery stated“it bothers me when I don’t earn anything without work.(…) In the coming four weeks I cannot work and that’s a shame because I need(money) to go abroad”(BEMF16). Although Ring Twice and Deliveroo provide insurance to workers, the limited coverage of the policy causes additional costs to accrue: “A customer claimed that I had drilled through the upholstery of his terrace, resulting in water seepage. It turned out that I wasn’t insured for that, so I had to reimburse him myself.”(BEMF28). Furthermore, workers on all platforms face limited control over their working time. Food delivery couriers feel obliged to deliver during lunch and dinner times, as they are unable to access enough orders at other times. Likewise, not knowing beforehand how long it will take to complete a task can lead to unpaid labour, for example in the case of unclear task descriptions like on Ring Twice: “I thought it would take two hours, but it took me four. The job add said:‘place a faucet’, but I found out that the faucet was inaccessible, I had to dismount two pieces of furniture to reach it.”(BEMF21) Freelancers and domestic workers often have to adapt their schedule to clients’ needs, for example when clients want them to complete tasks“within 24 hours”(BEMF41). This lack of control over their working schedule is underpinned by platforms’ rating systems, as failing to complete a task leads to a drop in workers’‘job success score’(Upwork) or ‘realisation rate’(Ring Twice). Finally, platforms constrain workers’ ability to enjoy and develop themselves through their work. Couriers express a desire to“be busy working outside”(BEMF39), however, they often find themselves compelled to“wait one hour in a restaurant”(BECM08) instead. Domestic workers affirm that their client relationships are strongly affected by the rating system:“We know that we will be evaluated right after” (BEMF24). This is why Ring Twice clients can ask workers to carry out additional unpaid tasks that workers feel obliged to complete to maintain a good rating. Similarly, newcomer Upworkers report“applying for everything I could potentially do”(BECM01) to increase their rating score. Consequently, they often end up completing many simple tasks unrelated to their profession: “I had to suppress the identifying data from an interview. I thought‘this is not a job for a translator’”(BECM02). COMMITTING UNPAID LABOUR AS AN INTENTIONAL ACTION AND CONSTRAINTS ON WORKERS Workers try to mitigate the constraints imposed by platforms by intentionally committing their unpaid labour. First of all, one practice adopted by workers on all three platforms is to forego income in the hope of accessing work, clients or higher pay. For Deliveroo couriers, this takes on the form of rejecting orders – and hence not being paid for them – in the hope of receiving an order with a higher fee or shorter waiting time. To this end, couriers engage in unpaid communications with restaurants to retrieve information on“how long(the food preparation) is going to take(…); if it takes 10 minutes or longer I cancel the order”(BEMF16). Couriers accept the unpaid time involved in waiting for the next order “because it still makes up for the time I would have waited in the(restaurant)”(BEMF35). Workers on Ring Twice and Upwork report trading in income for a higher rating, which is essential to accessing work: “They wanted to tip me because they were very satisfied. I asked them to write an evaluation instead”(BEMF25). By strategically“lowering my price to€ 6 to get the ratings and subsequently raising my price again”(BEMF31) – which often involves unpaid negotiations with clients – domestic workers and freelancers construe their unpaid labour as a process of“learning” to set the right price(BECM01; BEMF01) and“gaining confidence”(BEMF15) in their abilities as entrepreneurs. Moreover, they try to make their work more predictable by engaging in communications and exchanging information with clients. For example, some domestic workers“like to pay a non-paid visit to see what needs to be done and what(tools) I’ll need”(BEMF23) be5

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK fore performing the job. Food delivery and domestic workers sometimes invest money in new tools, bikes or scooters to work more efficiently, while some Upworkers decide to purchase a‘Freelancer Plus’ account, giving their profile more visibility and hence reducing exposure to unpaid job searching. However, these actions can be at odds with workers’ financial constraints, as some of them cannot afford to do unpaid or underpaid work: “two cents a word – you can’t live off that in Belgium(…) I have to find the limit I can handle, how much money I should make to survive”(BECM02). Even those workers who have a financial buffer sometimes refrain from engaging in unpaid labour because the returns are too low, given the already low pay on platforms: “€ 11/hour to rent an electric bike is really not worth it” (BEMF40) As a consequence, some workers decide to(temporarily) exit the platform and/or move to other jobs where payment is higher. Secondly, workers intentionally engage in unpaid labour by organising their life around platform work, thereby sacrificing free or private time. Seeking to obtain as many orders as possible, couriers make themselves available whenever they can, especially at lunch and during dinnertime. Similarly, domestic workers and freelancers report a need for constant availability involved in job searching and responding to client requests, and using their free time to complete“tasks immediately when I receive them on the weekend”(BEMF01). While couriers have little leeway to change their hours without losing income, freelancers are able to gain back some flexibility by securing“customers for life”(BEMF02) who regularly provide them with work. Nevertheless, all workers face private limits such as family responsibilities preventing them from working long or unsocial hours: appraisal”(BEMF04). Furthermore, domestic workers sometimes find their work more meaningful when they spend time on additional tasks or unpaid interactions with clients: “after completing a job I sit next to(the client) in the kitchen, we drink tea and talk for an hour(….); I enjoy having this social contact”(BEMF33). However, workers experience physical limits to their engagement with intensified work, which is why they sometimes reduce their working hours:“I never work all day(…) if you cycle intensively, after 4 hours you are exhausted”(BEMF37). While domestic workers can charge more for physically demanding jobs, this strategy is unavailable to couriers whose fees are set by Deliveroo. In awareness of their limited insurance coverage, both couriers and domestic workers sometimes work more slowly and carefully to decrease the risk of accidents. On the other hand, freelancers report mental exhaustion due to the continuous pressure to work and apply for jobs, which is why they sometimes limit their platform work: “with these platforms, you can suffer a burnout(…) if you don’t take some time off and set boundaries”(BECM01). Fourth, workers engage in self-promotional labour in order to access jobs and pay by improving their position on the platform. While Deliveroo couriers cannot distinguish themselves based on ratings, workers on Upwork and Ring Twice actively try to stand out from the competition by improving their rating score. They make distinctive, attention-grabbing job applications and craft attractive profiles in which they advertise their skills and experiences:“I highlight that I’m a teacher(…); I’m not just going to fix(your computer). I’m also going to explain what I’m doing”(BEMF24). Some freelancers report attending coaching sessions to learn more about how to promote themselves online. This helps them reduce their unpaid job search:“the clients are contacting me now instead of me contacting the clients”(BEMF15). “I can’t leave my youngest child alone at home in the evening”(BEMF31) Hence, some workers choose to limit their platform labour, as they“don’t want to neglect my social life”(BEMF23). Others state that being able to invest unpaid time in platform work is only possible because they are retired(e.g. BEMF29), temporarily unemployed(e.g. BEMF33), or students with a lot of time at their disposal(e.g. BEMF16). Third, workers intentionally intensify their work, tackle additional tasks or lower their pay in order to make their platform jobs more enjoyable or meaningful. Couriers engage in income-maximising‘games’ which consist of“cycling fast (…) to reach my goal of 3 orders per hour”(BEMF37). Similarly, freelancers and domestic workers sometimes consider their job applications as“a game of being the first one” (BEMF04). Working extra hard to deliver excellent work can give workers a feeling of“pride”(BEMF34). They perceive “client ratings as a source of satisfaction, motivation and Yet, this sometimes clashes with workers’ career and skill-development goals. On the one hand, workers perceive their self-promotional labour as repetitive and draining:“Applying is like doing the same thing every day, getting depressed, getting frustrated”(BECM01). On the other hand, workers experience that unpaid self-promotion does not help them to build a career outside the platform or to evolve towards more stable jobs: “I’ve been on Upwork for two years. I’ve completed almost 70 jobs. I’m still trying to find more stable work(…) every three days I have to find something else”(BECM02). This is why freelancers and domestic workers try to get repeat customers and sometimes take their client relationships outside the platform. Some leave the platform because they believe that they can receive more recognition for their skills elsewhere. 6

Workers’ experiences of unpaid platform labour as intentional action Table 1 Engagement with unpaid labour as intentional action and workers’ constraints Workers’ actions Foregoing income and investing money to access(higher) pay, work, or clients Organising life around platform work: foregoing free and private time to access or finish work Intensifying work, doing additional tasks, or charging less to make work more enjoyable or meaningful Standing out from the competition: self-promotion in order to access work and/or clients Source: own elaboration Corresponding engagement with unpaid labour Cancelling orders Trading in pay for a higher rating Communicating with restaurants and clients Lowering and then raising the price Price negotiations with clients Communications with clients Free visits to clients Free sample work Buying or renting bike / scooter / car / tools Purchasing paid account Unpaid commuting Constant availability Working unsocial hours(lunchtime or dinnertime, evenings/nights, weekends, holidays) Prolonging working hours to finish tasks to avoid non-payment Adapting schedule to client demands ‘Games’ involving working quickly ‘Games’ involving job applications Working hard to deliver excellent work Pursuing high rating score as a source of satisfaction Additional tasks and interactions with clients Charging less for enjoyable tasks and from clients in need Making distinctive profile and job applications: advertising skills and professional experiences Pursuing an excellent rating score: good reviews, completing many jobs Workers’ constraints Financial constraints: not able to afford unpaid labour, financial return is too low Private constraints: free time, private life, family responsibilities, etc. Physical and mental health constraints: exhaustion, risks of accidents, burn-out, overwork, etc. Career and skill development constraints: not able to build a career outside the platform A summary of workers’ intentional actions, their corresponding engagement with unpaid labour and the limits of their engagement can be found in table 1. 1.5 CONCLUSION We have investigated how and why platform workers engage in unpaid labour by considering the constraints imposed by platforms and workers’ intentional actions in a food delivery, domestic work and on an online freelancing platform. Our findings revealed the different significance workers attribute to their unpaid activities and the limits of their engagement. Specifically, we illustrate that workers can experience unpaid labour as a result of coercion by platforms, but that they also engage in it as an intentional activity. The intentional actions we shed light on show that workers’ engagement with unpaid labour means making creative use of their financial, private, physical, mental, skill and career-related resources as well as of the tools offered by platforms, such as rating systems. However, workers also limit their engagement when they are unable or unwilling to fall back on these resources. We find differences between platforms which are due to the different scope of discretionary latitude allowed (e.g. fixing workers’ pay(Deliveroo) versus workers setting prices themselves(Ring Twice)), but also due to the nature of the work. For example, freelancers can creatively deploy their professional skills and domestic workers can have prolonged face-to-face interactions with clients, while couriers lack these possibilities. The findings imply that studying workers’ experiences and relating them to their overall work and life situation is key to understanding why and under what conditions they engage in unpaid labour. While confirming that unpaid labour is part and parcel of platform work(Pulignano et al., 2022b), our research shows that platforms reducing costs and shifting risks to workers only partly explains the prevalence of unpaid labour. We move beyond assumptions under which unpaid labour is merely considered to be an outcome of platform control by illustrating that many unpaid activities rely on the initiative of workers who bring their own resources to platforms and actively attempt to make sense of their work. 1.6 ACKNOWLEDGEMENTS This study was developed within the framework of two interlinked research projects: ResPecTMe, funded by the European Research Council(ERC) under the European Union’s Horizon 2020 research and innovation programme(Grant Agreement number 833577), and‘Precarious work in the on‐line economy. A study on digital workers in Belgium and the Netherlands’, funded by the Flemish Research Council FWO(Flemish Research Council Project Number G073919N). 7

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK BIBLIOGRAPHY Anwar, M.A.,& Graham, M.(2020). Hidden transcripts of the gig economy: labour agency and the new art of resistance among African gig workers. Environment and Planning A: Economy and Space, 52(7), 1269–1291. Barratt, T., Goods, C.,& Veen, A.(2020).‘I’m my own boss…’: Active intermediation and‘entrepreneurial’ worker agency in the Australian gig-economy. Environment and Planning A: Economy and Space, 52(8), 1643–1661. Berg, J., Furrer, M., Harmon, E., Rani, M.,& Silberman, M.S.(2018). Digital labour platforms and the future of work: Towards decent work in the online world. Geneva: International Labour Office. Briziarelli, M.(2014). The ideological reproduction:(Free) labouring and (social) working within digital landscapes. tripleC: Communication, Capitalism& Critique. Open Access Journal for a Global Sustainable Information Society, 12(2), 620–631. Drahokoupil, J.,& Piasna, A.(2019). Work in the platform economy: Deliveroo riders in Belgium and the SMart arrangement. ETUI Research Paper-Working Paper No. 01. Brussels: ETUI aisbl. FOD Financiën.(2022, June 28). Deeleconomie – lijst van de erkende platformen. Retrieved at: https://financien.belgium.be/sites/default/files/ downloads/127-deeleconomie-lijst-erkende-platformen-20220628.pdf Fuchs, C.& Sevignani, S.(2013). What Is Digital Labour? What Is Digital Work? What’s Their Difference? And Why Do These Questions Matter for Understanding Social Media? tripleC: Communication, Capitalism& Critique, 11(2), 237–293. Gerber, C.& Krzywdzinski, M.(2019). Brave New Digital Work? New Forms of Performance Control in Crowdwork. In S.P. Vallas& A. Kovalainen(Eds.) Work and Labor in the Digital Age(Research in the Sociology of Work, Vol. 33),(pp. 121–143). Bingley: Emerald Publishing Limited. Gioia, D.A., Corley, K.G.& Hamilton, A.L.(2012). Seeking qualitative rigor in inductive research: Notes on the Gioia methodology. Organizational research methods, 16(1), 15–31. Glavin, P.& Schieman, S.(2022). Dependency and hardship in the gig economy: The mental health consequences of platform work. Socius, 8, 1–13. Holtum, J.P., Irannezhad, E., Marston, G.& Mahadevan, R.(2022). Business or pleasure? A comparison of migrant and non-migrant Uber drivers in Australia. Work, Employment and Society, 36(2), 290–309. Huws, U., Spencer, N.H.& Syrdal, D.S.(2018). Online, on call: The spread of digitally organised just‐in‐time working and its implications for standard employment models. New technology, work and employment, 33(2), 113–129. Kässi, O.,& Lehdonvirta, V.(2018). Online Labour Index: Measuring the Online Gig Economy for Policy and Research. Technological forecasting and social change, 137, 241–248. Maffie, M.D.(2022). Becoming a pirate: independence as an alternative to exit in the gig economy. British Journal of Industrial Relations, 1, 1–22. Moore, S.,& Newsome, K.(2018). Paying for free delivery: dependent self-employment as a measure of precarity in parcel delivery. Work, Employment and Society, 32(3), 475–492. Paelinck, G.(2020) Regeling rond onbelast bijverdienen vernietigd door Grondwettelijk Hof. VRT Nieuws. Available at: https://www.vrt.be/vrtnws/ nl/2020/04/23/bijverdienen/[Accessed 11th July 2022]. Pulignano, V.,& Franke, M.(2022). Control and consent regime dynamics within labour platforms. Work in the Global Economy, 1(aop), 1–27. Pulignano, V., Marà, C., Domecka, Franke, M.,& Muszyński, K. (2022a, forthcoming). Informality of employment in digital care platforms: A study on the individualization of risk and unpaid labour in mature market contexts. Pulignano, V., Grimshaw, D., Domecka, M., Vermeerbergen, L.,& Franke, M.(2022b, forthcoming). Unpaid Labour and Worker Performance on Digital Platforms: Why Differences in Socially Embedded Technologies and Worker Autonomy Still Matter. Pulignano, V.,& Marà, C.(2021). Working for Nothing in the Platform Economy. Forms and Institutional Contexts of Unpaid Labour. Solidar – Thematic Publication. Retrieved at: https://www.solidar.org/system/ downloads/attachments/000/001/410/original/Thematic_publication_ Working_for_Nothing.pdf?1638980551 Pulignano, V.,& Morgan, G.(2022). The‘Grey Zone’ at the Interface of Work and Home: Theorizing Adaptations Required by Precarious Work. Work, Employment and Society, 0(0). https://doi.org/10.1177/09500170 221122507 Rosenblat, A.,& Stark, L.(2016). Algorithmic labor and information asymmetries: A case study of Uber’s drivers. International Journal of Communication, 10, 27. Shapiro, A.(2020). Dynamic exploits: calculative asymmetries in the on‐ demand economy. New Technology, Work and Employment, 35(2), 162–177. Shevchuk, A., Strebkov, D.,& Tyulyupo, A.(2021). Always on across time zones: Invisible schedules in the online gig economy. New Technology, Work and Employment, 36(1), 94–113. Shibata, S.(2020). Gig work and the discourse of autonomy: Fictitious freedom in Japan’s digital economy. New Political Economy, 25(4), 535–551. Srnicek, N.(2016). Platform Capitalism. Cambridge: Wiley. Terranova, T.(2000) Free labor: producing culture for digital economy. Project Muse, 18, 33–58. Vallas, S.P.(2019). Platform capitalism: What’s at stake for workers?. New Labor Forum, 28(1), 48–59. Los Angeles, CA: SAGE Publications. Walker, M., Fleming, P.,& Berti, M.(2021).‘You can’t pick up a phone and talk to someone’: How algorithms function as biopower in the gig economy. Organization, 28(1), 26–43. Webster, J.(2016). Microworkers of the gig economy: Separate and precarious. New Labor Forum, 25(3), 56–64. Los Angeles, CA: SAGE Publications. Wood A., Graham, M., Lehdonvirta, V.,& Hjorth, I.(2019). Networked but Commodified: The(Dis)Embeddedness of Digital Labour in the Gig Economy. Sociology 53(5): 931–950. 8

Coalitional Power in the Digital Economy: Alliances of Gig and Tech Workers 2 COALITIONAL POWER IN THE DIGITAL ECONOMY: ALLIANCES OF GIG AND TECH WORKERS by Valentin Niebler(Humboldt University of Berlin, Institute for European Ethnology/ Berlin Institute for Migration Research) ABSTRACT Collective action in the tech industry has become a widely recognised phenomenon today. Low-paid gig workers have been at the forefront of these efforts, but more recently strikes and protests by higher-paid tech workers have taken place as well. This article investigates a case where both gig workers and tech workers have joined forces. Based on empirical data from Berlin, Germany, I analyse how gig workers and tech workers have generated‘coalitional power’ vis-à-vis a delivery tech company. I argue that, although more the exception than the rule, coalitions between gig workers and tech workers are possible, especially if both groups can cite common lines of conflict with capital and have a common legal context. In the Berlin case, the migration status of both gig workers and tech workers and the legal instrument of works councils has been a crucial requisite. ities of different layers within the workforces of these firms have become more apparent in recent years and are leading to increasingly overlapping conflicts of worker groups vis-avis management and owners of these companies. Based on analysis of preliminary fieldwork in Berlin, I describe possible elements of such alliances and the development of ‘common lines of conflict’ as pre-conditions for their success. The article starts out by laying out the context of labour struggles of both gig workers and tech workers, as well as the(scarce, but existent) history of efforts by both groups to form coalitions. After describing the methodology, the article provides a description and analysis of a coalition during a series of collective action campaigns at the company Gorillas regarding specific forms of worker power and a possible future of such labour alliances. 2.1 INTRODUCTION In July 2022, the delivery tech firm Lieferando organised a special after-work event for its employees in Berlin. A pool party was supposed to bring together the company’s staff and management with“food, drinks, and a pool, exclusively for you”(Lieferando Workers Collective, 2022). However, not the entire staff of the company was invited to the party. Lieferando delivery riders, although formally employed by the company, were explicitly excluded from the invitation. The scandal led to a public protest by riders in front of the venue, highlighting not only the tension between gig workers and management, but also between the higher and lower-paid staff groups of the company. The story of the Lieferando pool party exhibits a familiar narrative when it comes to the highly polarised tech industry. Well-paid, secure and well-appreciated professionals such as software engineers are placed on one side, and precarious and neglected gig workers on the other. Without denying the existence of such rifts, this article takes a somewhat different angle on the issue. Reflecting on organising efforts throughout the tech sector in the last years, the article looks at instances of coalition-building between gig workers and tech workers in the digital economy. I argue that vulnerabil2.2 BACKGROUND: LABOUR STRUGGLES IN GIG AND TECH The tech industry has been described as one of the key industries of contemporary capitalism(Zuboff, 2019; World Bank, 2019). I understand the tech industry as a network of corporations that derive their revenue from the employment of digital technology, most visibly in the form of digital platforms(Kenney/Zysman, 2020). 2 While it appears rather remarkable that labour conflicts have only recently become prominent, the industry has also developed many managerial and ideological methods to fragment labour relations in ways that conventional companies could not do or not do to the same extent(Barbrook& Cameron, 1996; Altenried 2022). Labour dynamics in tech are therefore often seen as blueprints for the future of work and management more 2 Examples of tech companies range from monopoly-seeking and transnational players such as Amazon, Google or Uber to smaller and more regional players. Most companies in the sector are currently characterised by vast amounts of(venture) capital investments and expectations of rapid growth. Companies in the tech industry make use of paid and unpaid labour in various segments, be it through software engineering, on-site platform work, web-based‘cloudwork’ or traditional forms of contingent labour(comp. Fuchs, 2012; Srnicek, 2017). 9

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK generally, although this remains a subject of debate(Ellmer et al., 2019; Schaupp, 2021; Azzellini et al., 2022). In the following, efforts at collective action by gig workers and tech workers will be described, while some tentative definitions of these groups will be proposed. 3 The term gig work has been established as an umbrella term for low-paid service jobs in the platform economy, in fields such as delivery, driving and cleaning as well as for solely web-based tasks such as content moderation or image recognition(Woodcock& Graham, 2020). In the European Union, around 2 percent of workers perform platform gig work as their main income source, totalling 4.3 million workers (Pesole et al., 2018). Gig workers often work remotely or mobile, as self-employed or subcontracted labourers and usually with little training or security. They are faced with a wide-ranging fragmentation in their work: from legal fragmentation through forms of sham self-employment, technological fragmentation through algorithmic management and all the way to spatial fragmentation through the diffusion of labour across cities or countries(Altenried et al., 2020; Della Porta et al., 2022: 10f). Despite such obstacles, gig workers in tech firms around the world have organised collectively in recent years, even leading to a surge in labour struggles generally speaking(Bessa et al., 2022). Especially labour struggles in delivery and ride hailing, often by independent worker collectives, have become notorious and have challenged the leverage of companies, leading to company exits, corporate losses and legislative interventions(Vandaele, 2018). The term tech work describes occupations such as software engineers, technical writers, UX designers and other white-collar staff employed at tech companies(Dorschel, 2022). Tech workers are wage-dependent employees who work predominantly on cognitive tasks, earn middle- to higher-level salaries and often possess a(relative) secure employment status. In 2020, tech workers accounted for 4.6 percent of total employment in the European Union, totalling around 9 million workers(Rothstein, 2022). While many tech workers are employed in standard labour relationships, a considerable number also work as temp workers or contractors. According to Tarnoff, workplace conflicts involving tech workers appear to fall into three categories: issues revolving around wages and working conditions, concerns for safe and equitable workplaces, and discontent about the social harms of company products(Tarnoff, 2019). In addition to this, there is a volatile and venture capital-driven corporate environment, which is prone to job losses and fundamentally opposed to collective bargaining and trade unions. Since around 2016, tech workers have increasingly organised in the industry, inter alia through public walkouts and unionisation campaigns at companies such as Alphabet and Activision(Jaffe, 2021). In China, employees have for 3 It should be noted that my use of the term tech work differs from other authors(such as Tarnoff/ Weigel 2020), who use it to describe all forms of work within a tech company. I distinguish between tech work for higher-paid office work, and gig work for low-paid and technologically mediated work. Just as with the term tech company, definitional boundaries are not always clear and remain blurry. several years been organising against exhausting working hours(Tan& Weigel, 2022). Although collective action efforts by gig and tech workers usually evolve separately, they often deal with similar and related issues. They are faced with companies that are often highly volatile and short-lived, are subject to algorithmic control(especially gig workers) and confronted with ideological efforts to treat their work as play or leisure(especially tech workers). Given these circumstances and the vast concentration of power in the sector over the past decade, there has been debate about whether the various groups in the industry can join forces(Weigel, 2017). Along the lines of this debate, my article aims to explore the possibilities of coalitional work between tech and gig workers. So far, common efforts between higher-paid and lower-paid workers in tech are not ubiquitous. Given their different positions in the value chains of tech, lack of sectoral collective bargaining, and the fragmented terrain for organising work described above, this does not appear surprising. Some notable exceptions exist, however. Among the most wellknown is the project Turkopticon, a software tool that helps workers on the crowdsourcing platform Mechanical Turk to confront lack of transparency by collectively rating their clients(Silverman& Irani, 2016). The tool was developed by two software designers who are aware of working conditions and was eventually put into operation collectively with platform workers(ibid.). More recent cases include a series of workplace solidarity actions at Facebook in Silicon Valley and Amazon in Seattle, where tech workers have actively supported cafeteria workers and security guards in their unionisation efforts(Tarnoff, 2019; Weigel, 2017). Through the Alphabet Workers Union(AWU) at Google, it was possible to achieve more security for subcontracted data center personnel(Jaffe, 2021). If studied in more detail, cases of worker cooperation do not just offer insight into the strategic development of worker power. They also provide insight into class relations in the field. What specific circumstances enable cross-class solidarity in tech? How do different groups and actors relate to each other? What kind of actions are taken by protagonists, and how is power maintained in these cases? To shed more light on this, this article will in the following relate a current case involving such a labour alliance in Germany. The aim of the article is to expand knowledge and awareness of such cooperation and to provide conceptual footing for a systematic analysis of these phenomena. 2.3 METHODOLOGICAL APPROACH The analysis performed in this article focuses on efforts to form coalitions by workers during a series of labour struggles at the company Gorillas in Berlin. Empirical research was conducted between May 2021 and November 2022 and included nine qualitative interviews with workers and activists in Berlin as well as three expert interviews with trade union representatives and organisers. Ethnographic fieldwork from organising events, protests and three court hearings in Berlin 10

Coalitional Power in the Digital Economy: Alliances of Gig and Tech Workers were included in the research. The research was complemented by insight from a series of public discussion events including tech and gig workers on the issue. 4 Adopting an inductive approach, I first addressed events in the field and conducted interviews, and then developed analytical categories with which to frame the analysis of the material. Coding of the material was based on the qualitative analysis framework developed by Kuckartz(2010). To carry out a concrete analysis of worker power, my article employs the Power Resource Approach(PRA). The PRA is a heuristic research tool used to analyse the potentials of trade unions and social movements, and was developed in the context of trade union revitalisation research(Dörre& Schmalz, 2014). Based on earlier concepts of workers’ power resources(Wright, 2000; Silver, 2003), the approach builds on the premise that workers possess strategic choices in their conflicts with capital, depending on their position in the labour process, the labour market, their institutional context and their position in society(Schmalz et al., 2018). Analytically, the approach divides the potential of workers to act in their own interest into several pillars: structural power, associational power, institutional power and societal power. As part of societal power, the framework for coalitional power relates to“pursuing common goals and entering into mutual commitments”(ibid.) with other groups in society. Coalitional power is able to improve worker power by“harnessing the resources of other players[…] to mobilise support from these actors”(ibid.). With an extended notion of coalitional power, my analysis seeks to highlight how the pooling of power in the tech industry becomes possible, which obstacles it faces and what advantages it holds out for workers’ groups. 2.4 RESULTS Although Berlin ranks among the smaller hubs in the global tech industry, the city’s influence and concentration of investment capital has been growing steeply over the last decade (Staab, 2019: 95). Berlin attracts a mixture of domestic and global tech firms, which can draw from an increasingly international and often skilled pool of workers. A prominent case of both economic growth and labour conflict has been the startup Gorillas, a delivery firm specialising in rapid delivery of groceries. The company was founded in Berlin during the Covid-19 pandemic and expanded rapidly across Europe and beyond. Gorillas was able to raise over USD 1 billion in investments in only nine months, making it one of the highest valued companies in Europe’s startup scene(Ewen et al., 2022; Keane, 2021). At the height of its market power in 2021, the company employed over 1500 warehouse pickers and delivery riders in the city and around 600 employees at the company’s headquarters(Sell, 2021; Frank, 2022). With some exceptions, gig workers at Gorillas are formally employed by the company, which makes it possible for workers to invoke the co-determination instruments of German labour law. Conflicts around working conditions at Gorillas arose soon after the company started operating in Berlin, culminating in spontaneous strikes by gig workers in early 2021. Riders on bikes and pickers in warehouses complained about late and incorrect payments, as well as security and hygiene hazards (Ewen et al., 2022). A group of riders founded a workers collective, which later developed into the Gorillas Workers Collective(GWC). In order to establish a formal representation of workers in the company, the group reached out to several groups and unions in Berlin. Among them was the Berlin Tech Worker Coalition(TWC), a collective of organised tech workers in Berlin. Since their foundation in 2019, the Berlin TWC has been involved in several forms of labor organising. The collective offers training workshops for workers in tech, mostly relating to the establishment of works councils. Works councils are institutions of workers’ representation at the company, are not tied to unions and are protected by labour law, and have become an increasingly strategic tool for workers and activists in Germany’s gig economy. Alongside with other actors, GWC and TWC met in the spring of 2021 at several meetings to discuss strategic steps towards founding a works council. As one TWC members recalls:“It was around March in 2021, Gorillas reached out to us.[…] They were asking what kind of concrete support we can give them, trainings or what next steps there are” (TWC_B01). Through their meetings, TWC members became more aware of the issues of Gorillas workers and got more involved with helping them against union-busting efforts. This became even more important when the conflict with the company escalated a few weeks later, and a series of wildcat strikes and blockades against the company started. Members of TWC then joined in the mobilisation and protests. Overall, the coalition efforts of GWC and TWC can be broken down into three main groups. First, passing on knowledge and resources, secondly shielding off risk, and thirdly amplifying voices. First of all, TWC members helped to facilitate the resource-intensive process of initiating the formation of a works council. 5 They did so by securing meeting spaces (making sure management could not enter the space), offering directions and organising translators to ensure participation of all workers:“Concretely, we helped them finding translators, provided security with entrance policies, also just/ people standing in the hallways to help with directions in the hotel building, and stuff like that“(TWC_B01). For this, the group was able to make use of their existing networks in the city. Part of the support appears to involved explaining the complicated details and provisions of works council 4 The event series“Challenging Tech” took place online in May and June 2021, organised by the Centre for Emancipatory Technology Studies and Rosa Luxemburg Foundation. Documentation of the events can be accessed at: https://emancipatory.technology/news/ documentation-of-the-discussion-series-challenging-tech 5 The very formalised works council format necessitates an initiation process that requires workers to vote on an electoral commission(cf. Fitting et al., 2022). To establish this electoral commission, workers meet in a specific place and elect the members of the commission. As this process requires a lot of resources and knowledge, it is usually hard for people not familiar with the issue to pursue. This is especially the case with migrant gig workers. 11

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK legislation in English and other languages. In particular, TWC could contribute experience with organising in fast-paced tech companies, which tend to differ in their approach toward conventional companies. Secondly, TWC members also engaged in direct collective action together with the GWC in a series of blockades and wildcat strikes that occurred as the conflict escalated further during the month of June 2021. Strikes and blockades took place in front of the company’s hyperlocal warehouses, which deliver groceries to households in a neighborhood. Workers and activists blocked the doors of warehouses and turned bikes upside down so they could not be used anymore. According to TWC members, a main objective of their involvement in the actions was to shield workers from risks and exposure in the public arena. As actions were publicly recorded, protesting Gorillas workers were at risk of being sanctioned for their actions. In some cases, TWC members were also involved in negotiations with warehouse management staff. Later on, TWC members were involved in supporting GWC workers in cases before the labour courts. Thirdly, TWC supported the GWC by helping with press work, giving interviews and directing international attention tow the cause. Concretely, tech workers amplified the voice of Gorillas workers at a time when attention for their cause was still limited. With reference to their help with press work, one TWC member recalls:“[…] that was something that the GWC was really overwhelmed with, the number of journalists who asked everything from really basic questions […] helping with copy-editing draft releases was one thing, and then just fielding phone calls, having chats, here is this network, can someone contact them?”(TWC_B01). Most clearly, TWC helped draw international attention to the case through their well-established networks in social media and to other actors in Germany and beyond. Looking at all three forms of support, it can be said that workers at TWC were in the possession of experience, contacts and resources that the GWC was lacking at the time. As employees at several firms, they were confronted with a similar context of legal tools, specifically the works council. This made cooperation likely and useful in practical terms. Asked what TWC had been helpful with, a GWC member recalls: I think they gave us[…] this internal structure that we needed[…][and] access to the spaces we needed. They helped us with twitter,[…] helping us go viral. Setting up a Soli-Fund, setting up e-mails, a lot of things. Not just that, but they also showed up.[…] Like they showed up to protests, demonstrations, strikes”(GWC_B03). Throughout the research, both tech and gig workers referred to their role as migrant workers, which most of them identified as a common challenge. Regarding the difficulties of organising, a TWC interviewee stated that“what makes it hard to organise is[…] a highly migrantised workforce. Specifically in Berlin, you have a group who are not so familiar with the German[…] legal tradition, or speaking German.” (TWC_B01) Similar aspects were raised by gig worker interviewees. This suggests that their experiences as recently arrived migrant workers(namely language difficulties, legal vulnerability, racism and visa issues) likely shaped their experience in organising the workplace and served as a bond during the cooperation. This also became visible in rifts and conflicts with the trade union ver.di, which according to the interviewees failed to work together with Gorillas workers due to conflicts over what tactics to use(the union opposed wildcat strikes, which it legally cannot support) and due to condescending comments made by a union representative about the German language skills of the workers. While tech workers faced less open hostility from unions, many shared concerns and ambivalent experiences. The vulnerabilities tied to their migrant status varied across individuals and were less severe for most tech workers, but appeared similar enough to relate to each other to some extent. Although far from all demands in the Gorillas conflict could be met, some were indeed satisfied. A works council could be initiated despite heavy union-busting efforts by Gorillas, the company had to step up safety measures for workers due to public pressure, and developments towards labour law reform were written into the coalitional agreement of Germany’s new federal government in the fall of 2021(Bund Verlag, 2022). All of these changes are very likely connected to the protests and the high degree of public pressure they created on the company and on policymakers. Just like cooperations with other groups, the gig-tech worker coalition contributed to the establishment of longer-lasting networks in the city and beyond. Looking at the case through the lens of power resources makes it possible to reflect on the pooling of power by both groups. Generally, gig workers at Gorillas appear to have generated associational power(the ability to mobilise workers) very quickly, meaning they have been able to come together collectively to go on strike at their workplaces. 6 However, Gorillas workers lacked the knowledge and experience to create a works council, a tool of institutional power(the ability to make use of institutionalised rights) that the TWC as well as other groups could assist them with through workshops and knowledge-sharing. The cooperation helped to establish the works council as an important, legally secure tool of formal workers’ representation and decision-making at the company. The support of tech workers during wildcat strikes points to two additional aspects: on the one hand, an increase in associational power through direct support, and on the other, a boost in societal power(the ability to receive support by the general public) that was established by the ‘amplifying of voices’ on social media and traditional news outlets described above. By boosting the message of Gorillas workers through well established connections and accounts, TWC was able to introduce the GWC as a legitimate actor vis-à-vis the company in the public arena. Overall, the cooperation increased the strength of the GWC and made it possible to achieve some workers’ goals. This 6 This is a potential that specifically grocery delivery couriers and pickers appear to have, as they are located in physical workspaces and can socialise and create common bonds there easily. 12

Coalitional Power in the Digital Economy: Alliances of Gig and Tech Workers is also remarkable because the structural power(the ability to withdraw labour power) of the tech workers and other cooperation partners did not play a role here. The TWC members involved were not part of Gorillas staff, but could nevertheless reinforce the effort, most likely with their own experience of union-busting techniques at tech companies. The fact that no Gorillas tech workers were involved in the actions also points to the advantages and disadvantages of in-house alliances. On the one hand, in-house tech workers have structural power and essential knowledge about workplace and management structures. On the other hand, they are often not trusted actors and might have conflicting interests or loyalties within the company. Therefore, cooperation with non-affiliated tech workers might make more sense for gig workers, especially if in-house staff is not unionised. Lastly, some questions remain on motivations for the described alliance. Did the TWC also benefit from the alliance, or was its support mainly altruistic? Some interviews suggest that the involvement in gig worker struggles helped TWC gain legitimacy among other groups who often view tech workers as‘part of the problem’ in labour conflicts. Again, this points to an increase in coalitional power for the group. Still, the reasons for the involvement of TWC as an advocacy group beyond a single company are naturally different. As with other activist groups, cooperation might also evolve from notions of“groundless solidarity”(Elam, 1994) or at least without expectation of reciprocity(Jaeggi, 2021). Never­theless, familiarity with the industry and an overall sense of fighting a common enemy have likely contributed to the cooperation. Generally, the strikes at Gorillas were the product of a large number of coalitional efforts, among which there was only one with TWC. Crucial and often closer cooperation took place with the syndicalist base union FAU, other rider collectives, the anti-fascist migrant movement Migrantifa, as well as a media collective and several individual activists and lawyers. and its overall increase through cooperation. Although just a small part of the Power Resource Approach, coalitional power as a basis might be useful in developing more elaborate concepts with which to analyse and compare specific forms of coalitions, such as those at Amazon Mechanical Turk, Facebook and Alphabet. The analysis also suggests that traditional trade unions need to find ways to respond to such conflicts and especially to the demands and tactics of precarious migrant workers. The main part of the observed coalition between TWC, GWC and other groups appears to have filled a void that formal trade unions failed to provide. Although some experiments with semi-formal and informal unionism have been made in the last years(Heiland, 2020; Niebler& Kern, 2020; Basualdo et al., 2021), unions in Germany and Europe have so far been somewhat reluctant to utilise the tactics of migrant and gig worker organising. Considering the most recent crisis of the tech industry and its lay-offs across divisions, opportunities for conflict and organising will surely arise. In the long run, cross-class coalitions could be an important lever in winning back worker power. 2.6 ACKNOWLEDGEMENTS I am grateful for very helpful comments received from Franziska Baum, Florian Butollo, Dalia Gebrial and Charalambos Tsekeris at the PhD Summer School“Trajectories of Platform Capitalism and Platform Work”, hosted by Friedrich-EbertStiftung and Weizenbaum Institute in Berlin in September 2022. 2.5 CONCLUSION Although often at the opposite ends of the corporate hierarchy, gig workers and tech workers are able to enter into labour alliances and thereby advance their worker power. During the labour conflict at Gorillas in Berlin, organised tech workers helped gig workers by sharing their resources and networks, shielding off risks during direct action and amplifying gig workers’ voices in public. Shared or related experiences and challenges as migrant workers and migrant organisers provided common‘lines of conflict’ for the two groups, both vis-à-vis the corporation and to some extent in contrast to traditional trade unions. Shared experiences of migrant workers in tech appear to be a fruitful avenue for gig-tech alliances more generally, as the share of migrant labour within both gig and tech work professions is high in many countries(Amrute, 2016; Altenried, 2021). A second common issue is the legal context, which made the instrument of the works council relevant for both groups. The concept of coalitional power has proven useful as a lens through which to look at the combination of power resources 13

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK BIBLIOGRAPHY Altenried, M.(2021). Mobile workers, contingent labour: Migration, the gig economy and the multiplication of labour. Environment and Planning A: Economy and Space, 0(0) Altenried, M.(2022). The digital factory. The human labor of automation. Chicago, London: The University of Chicago Press. Jaeggi, R.(2021). Solidarität und Gleichgültigkeit. In: Susemichel, L., Kastner, J.(Eds.): Unbedingte Solidarität. Münster: Unrast. Jaffe, S.(2021). Organizing Big Tech. Rosa-Luxemburg Foundation, New York Office. Available online at https://rosalux.nyc/wp-content/uploads/ 2021/05/RLS-NYC-Organizing-Big-Tech.pdf. Altenried, M., Niebler, V.& Macannuco, J.(2020). Platform Labour: Contingent Histories and New Technologies. Soft Power, 7(1), 255–265. Amrute, S.(2016). Encoding race, encoding class. Indian IT workers in Berlin. Durham, London: Duke University Press. Basualdo, V., Dias, H., Herberg, M., Schmalz, S., Serrano, M.& Vandaele, K.(2021). Building workers’ power in digital capitalism. Old and new labour struggles. Bonn: Friedrich-Ebert-Stiftung(Trade Unions in Transition 4.0). Barbrook, R.& Cameron, A.(1996). The Californian ideology. Science as Culture 6(1), pp. 44–72. Bessa, I., Joyce, S., Neumann, D., Stuart, M., Trappmann, V.& Umney, C.(2022). A global analysis of worker protest in digital labour platforms. Geneva, Switzerland: International Labour Organization(ILO working paper/International Labour Organization, 70(July 2022)). Bund Verlag(2022). Große Erwartungen: Koalition will Mitbestimmung ausbauen. Available online at https://www.bund-verlag.de/aktuelles~ Grosse-Erwartungen-Koalition-will-Mitbestimmung-ausbauen~.html Della Porta, D., Chesta, R.& Cini, L.(2022). Mobilizing against the odds. Solidarity in action in the platform economy. In Berliner Journal für Soziologie 32(2), pp. 213–241. Dorschel, R.(2022). A new middle-class fraction with a distinct subjectivity: Tech workers and the transformation of the entrepreneurial self. The Sociological Review, 70(6), 1302–1320. Keane, J.(2021). Berlin’s Gorillas Raises$1 Billion Led By Delivery Hero. In Forbes, 10/19/2021. Available online at https://www.forbes.com/sites/ jonathankeane/2021/10/19/berlins-gorillas-raises-1-billion-led-by-deliveryhero/, checked on 4/22/2022. Kenney, M.& Zysman, C.(2020) The Platform Economy: Restructuring the Space of Capitalist Accumulation, Cambridge Journal of Regions, Economy and Society, 13(1): 55–76. Kuckartz, U.(2010): Einführung in die computergestützte Analyse qualitativer Daten. Wiesbaden: VS Verlag für Sozialwissenschaften. Lieferando Workers Collective(2022):“#Lieferando Summer Pool Party 2022? Rider und Driver sind explizit ausgeladen. Available online at https://twitter.com/LWC_Berlin/status/1541396332996354050/photo/2, updated on 8/1/2022, checked on 8/1/2022. Niebler, V.& Kern, A.(2020). Organizing YouTube: a novel case of platform worker organising. Trade Unions in Transition 4.0, Friedrich-E­ bertFoundation Pesole, A., Urzí Brancati, M.C, Fernández-Macías, E., Biagi, F.& González Vázquez, I.(2018). Platform Workers in Europe, EUR 29275 EN, Publications Office of the European Union, Luxembourg Rothstein, S.(2022). Recoding Power. Tactics for mobilizing tech workers: Oxford University Press. Silver, Beverly J.(2003). Forces of labor. Workers’ movements and globalization since 1870. Cambridge: Cambridge University Press. Dörre, K.& Schmalz, S.(2014). The power resource approach: An instrument to analyze trade union action capabilites. Industrielle Beziehungen 21(3), pp. 217–237. Schmalz, S., Ludwig, C.& Webster, E.(2018). The Power Resources Approach: Developments and Challenges. Global Labor Journal 9(2). DOI: 10.15173/glj.v9i2.3569. Elam, D.(1994). Feminism and deconstruction. London: Routledge. Ellmer, M., Herr, B., Klaus, D.& Gegenhuber, T.(2019). Platform workers centre stage! Taking stock of current debates and approaches for improving the conditions of platform work in Europa. Working Paper. Hans-Böckler-Stiftung. Düsseldorf. Ewen, J., Heiland, H.& Seeliger, M.(2022). Dynamiken autonomer Arbeitskonflikte im digitalen Kapitalismus: der Fall“Gorillas”. Schriftenreihe Institut Arbeit und Wirtschaft, No. 33/2022, Institut Arbeit und Wirtschaft(IAW), Universität Bremen und Arbeitnehmerkammer Bremen, Bremen Fitting, K., Engels, G., Auffarth, F.& Kaiser, H.(2022). Betriebsverfassungsgesetz. Handkommentar. 31. Auflage. München: Verlag Franz Vahlen. Fuchs, C.(2013). Class and Exploitation on the Internet. In Scholz, T.(Ed.). Digital labor. The Internet as playground and factory. New York: Routledge, pp. 211–224. Frank, M.(2022). Massenentlassungen bei Gorillas: Klage wird mitgeliefert. Available online at https://taz.de/Massenentlassungen-beiGorillas/!5856794/ Frege, C., Heery, E.& Turner, L.(2010). The New Solidarity? Trade Union Coalition-Building in Five Countries. In Frege, C., Kelly, J.,(Eds.): Varieties of unionism. Strategies for union revitalization in a globalizing economy. Oxford: Oxford Univ. Press, pp. 137–158. Schaupp, S.(2021). Technopolitik von unten. Algorithmische Arbeitssteuerung und kybernetische Proletarisierung: Matthes& Seitz. Sell, S.(2021). Neues aus der Lieferbotengesellschaft: Die Gorillas Worker wollen einen Betriebsrat und bekommen Hilfestellung vom Arbeitsgericht. Available online at https://aktuelle-sozialpolitik.de/2021/11/17/neues-ausder-lieferbotengesellschaft/ Srnicek, N.(2017). Platform capitalism. Cambridge, UK, Malden, MA: Polity(Theory redux). Staab, P.(2019). Digitaler Kapitalismus. Markt und Herrschaft in der Ökonomie der Unknappheit. Berlin: Suhrkamp. Tan, J. S.& Weigel, M.(2022). Organizing in(and against) a New Cold War: The Case of 996.ICU. Available online at https://direct.mit.edu/books/ oa-edited-volume/5319/chapter/3800164/Organizing-in-and-against-aNew-Cold-War-The-Case. Tarnoff, B.(2020). The Making of the Tech Worker Movement. In Logic Magazine, 5/9/2020. Available online at https://logicmag.io/the-making-ofthe-tech-worker-movement/full-text/ Tarnoff, B.& Weigel, M.(2020). Voices from the Valley: Tech Workers Talk About What They Do – and How They Do It: FSG Originals. Vandaele, K.(2018). Will trade unions survive in the platform economy? Emerging patterns of platform workers’ collective voice and representation in Europe. ETUI, Brussels. Irani, L.& Silberman, M.S.(2016). Stories We Tell About Labor. In Jofish K.(Ed.). Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems, pp. 4573–4586. Weigel, M.(2017). Coders of the world, unite: can Silicon Valley workers curb the power of Big Tech? In The Guardian, 10/31/2017. Available online 14

Coalitional Power in the Digital Economy: Alliances of Gig and Tech Workers at https://www.theguardian.com/news/2017/oct/31/coders-of-the-worldunite-can-silicon-valley-workers-curb-the-power-of-big-tech Woodcock, J.& Graham, M.(2020). The gig economy. A critical introduction. Cambridge, Medford, MA: Polity. World Bank(2019). World Development Report 2019. The Changing Nature of Work. Washington, D.C: The World Bank(World Development Report). Wright, E.O.(2000). Working-Class Power, Capitalist-Class Interests, and Class Compromise. American Journal of Sociology 105(4), pp. 957–1002. DOI: 10.1086/210397. Zuboff, S.(2019). The age of surveillance capitalism. The fight for a human future at the new frontier of power. New York: PublicAffairs. 15

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK 3 PLATFORM WORK IN MADRID, MILAN, AND SAN FRANCISCO: EXPLAINING MUNICIPAL RESPONSES THROUGH A MULTI-LEVEL GOVERNANCE FRAMEWORK by Maximilian Kriz(University of Glasgow, Department of Urban Studies) ABSTRACT Determining the factors that shape the governance responses of city governments to platform work and digital labour platforms constitutes one objective of scholars studying platform urbanism. Thus far, however, no analytical framework has been proposed which could illuminate the influences of state and non-state actors within institutional constraints on municipal responses. This paper seeks to fill that gap by mobilising multi-level governance. By subjecting qualitative data collected in Madrid(Spain), Milan(Italy), and San Francisco(USA) to four indicators of multi-level governance, the concept’s value in accounting for the influences behind each respective municipal response is demonstrated. Multi-level governance reveals the significant influences of national governmental hierarchies and municipal officials’ perception of platform work which, combined with non-state actors’ engagement, affect the distinct municipal response in each case. Furthermore, it highlights the crucial role municipalities can play, either in mitigating platform workers’ precarity or in promoting such work in cities, despite constrained competences or opposition from non-state actors. 3.1 INTRODUCTION The precarity of platform workers remains a worldwide challenge for policymakers and scholars alike. Due to the reliance of geographically tethered platforms on urban infrastructures and workforces(Caprotti et al., 2022), city governments find themselves at the centre of tensions created by platform work and manifested in the form of protests(Bessa et al., 2022), precarious work conditions(Prassl, 2018), and court cases(Cherry& Aloisi, 2017). In this connection, this paper addresses a question that has hitherto not been explored sufficiently: why do city governments respond to appbased platform work in different ways? In other words, why do they enact(or not enact) certain regulations to govern platforms using independent contractors who lack access to social protection, regular pay, and collective bargaining (Hooker& Antonucci, 2022)? Previous research, though generating highly insightful findings, has not produced an analytical framework that could supply satisfying answers. The framework should allow researchers to investigate empirical data, study the factors explaining municipal responses to platform work, and compare different cases. Proposing such a framework is the objective of this text. The tendency towards categorising urban-level regulation of platform work, instead of focusing on the determinants of municipal responses, has come at the cost of explanatory power and fails to explain why certain city governments are reluctant to support app-based workers. This creates a need for an analytical framework that provides a more holistic picture of why city governments react to platform work differently, while taking into account national governmental structures as well as the influence of non-state actors. The paper proposes the use of multi-level governance(MLG) to meet these requirements, and it proposes operationalisation of the concept by means of four indicators. These indicators are: 1. roles, views, and involvement of non-state actors; 2. the relationship between governmental layers regarding platform work; 3. the availability of competences and instruments at municipal level; and 4. the involvement of the municipality in the policy issue of platform work. Subjecting empirical data – collected through semi-structured interviews and policy papers in Madrid, Milan, and San Francisco – to the MLG indicators stresses the importance of governmental hierarchies and municipal officials’ perception of platform work in explaining each case’s distinct governance of the phenomenon. 3.2 BACKGROUND VARIED MUNICIPAL RESPONSES TO PLATFORM WORK The connection between platforms, platform work, and urban areas has by now been widely recognised, particularly by proponents of“platform urbanism”(Barns, 2014; Sadowski, 2020). Responses by governments to the expansion of platform work vary widely across cities, however. A growing number of academic articles have investigated how platform work is treated by different municipalities. Zanatta 16

PLATFORM WORK IN MADRID, MILAN, AND SAN FRANCISCO & Kira(2018), for example, traced the response to Uber’s arrival by São Paulo’s city government, stressing the political struggles that shape regulatory action in the Brazilian metropolis. Similar research projects in North America have revealed the potential steering power of municipal officials by concentrating on tensions and interests within city governments: Brail(2018) found that the municipality of Toronto acted with the aim of promoting innovation, while Flores& Rayle(2017) explained the approach of San Francisco’s City Hall to ride-hailing services as a result of single political actors favouring such platform services. Meanwhile, other studies have adopted a more comparative view by analysing several cities simultaneously, with a greater emphasis on metropolitan areas in the Global North. For example, Aguilera et al.’s(2021) study of municipal regulation of short-term rental platforms in Barcelona, Milan, and Paris showed how social struggles and local alliances of stakeholders make platforms a political issue. Yet, similar to Ardura Urquiaga et al.’s(2019) legal analysis and quantitative evaluation of rental app regulation in Barcelona and Madrid, it did not emphasise the issue of platform work. In contrast, Beer et al.(2017) and Wolf(2022) both focused on the regulatory reception of ride-hailing platforms in US cities, employing qualitative comparison and multivariate modelling, respectively, but did not venture any explanation for the different municipal approaches. Studying responses in three Canadian cities utilising a governance approach, Tabascio& Brail(2021) highlighted the key interplay between regional and metropolitan governments in explaining why, in contrast to Vancouver and Montreal, the municipality of Toronto regards ride-hailing platforms as a municipal responsibility. Other research projects have adopted a focus on governance similar to Tabascio& Brail in order to shed light on the varying municipal responses to platforms, albeit on a larger scale. Voytenko Palgan et al.’s(2021) paper on the municipal governance of the sharing economy in European and North American cities identified five mechanisms through which municipalities govern the sharing economy. Yet, their broad focus on the sharing economy, which went beyond platform work, has left nuances to be explored between reactions to different types of platforms and app-based work. Such a broad focus made Vith et al.’s(2019) research on urban responses to the sharing economy in 16 global cities less useful for understanding platform work in cities as well, although their investigation suggested that the framing of platforms – as proposed by Thelen(2018) for the national level – also seemed to be a decisive factor in regulation at the urban level. Thelen’s analysis challenged the assumption that comparable stakeholders in the governance of platforms hold the same interests across different locations, thereby underscoring the need for careful examination of local actors and their motives. Vidal et al.(Morell, 2018) adopted a similar approach to Vith et al. by classifying municipal governance modes of the sharing economy. The greatest contribution by these studies – from Tabascio& Brail to Vidal et al. – has been their focus on governance rather than solely regulation, lending more weight to the roles of municipalities as monitors or enablers of platform work. As Da Cruz et al.(2018) pointed out, the increasing number and diversity of actors involved in policymaking processes, in addition to a growing expectation of cities to become more self-reliant, justifies a more thorough examination of governance structures, how municipalities manage other actors and responsibilities, and with what effect. Previous articles have answered the questions of the‘what?’ and‘how?’ very effectively, proposing useful classifications for types of municipal governance platforms, but rarely with regard to the‘why?’. Furthermore, whenever studies address explanatory factors, they often do so by focussing too broadly on the sharing economy or outside any replicable framework that can identify the various potential drivers of municipal responses to app-based work. The ideal analytical framework hence needs to be able to account for city-­ specific governmental structures as well as non-state actors’ influence, including workers’ representatives or platform managers. Pressure from non-state actors at various governmental levels can matter as much as competences and available policy instruments. 3.3 THEORY AND METHODOLOGY ANALYSING MUNICIPAL POLICY CONTEXTS THROUGH MULTI-LEVEL GOVERNANCE The term multi-level governance(MLG) was first coined by Marks(1993: 392) and defined as“a system of continuous negotiation among nested governments at several territorial tiers”. Bache& Flinders(2004: 3) refined this definition and established that multi-level refers to the“interdependence of governments operating at different territorial levels”, while governance points to the“growing interdependence between governments and non-governmental actors”. MLG thus goes beyond intergovernmental relations and crucially “factor[s] in the participation of non-governmental actors” (Piattoni in Ongaro, 2015: 326). In other words, MLG can act as a“framework for interpreting governance in complex polities”(Ongaro, 2015: 2). To exercise this analytical function, MLG has been widely conceptualised along two axes, a vertical and a horizontal one. Thus conceptualised, the framework accounts for the distribution of government authority or competence vertically to state actors at other territorial levels, as well as horizontally to non-state actors (Bache& Flinders, 2004). The conceptualisation of MLG along two axes can convey a simplistic image about the policy processes it tries to explain, however. On one hand, non-state actors need not be confined to a single horizontal level. Such multi-level interaction can be the product of necessity when officials of one governmental level are unwilling to reciprocate interest groups’ willingness to engage, leading the latter to resort to another level. Furthermore, evidence gathered by Wolf(2022) demonstrated that platforms deliberately seek to influence higher levels to pre-empt regulation by municipalities. Conceptualising the role of non-state actors through the horizontal axis should therefore not create an illusion of equally powerful stakeholders located at a single governmental level. 17

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK On the other hand, the vertical axis can equally misrepresent the formal and informal relationship between governmental layers. The relationship is not necessarily hierarchical, but can take more interdependent or even independent forms (Bache& Flinders, 2004). Crucially, MLG alerts one to the limits and influences on policymakers embedded in a complex multi-level, multi-stakeholder environment. Its analytical value is thus particularly great in urban contexts. As Bramwell(2020) observed, city governments operate in a multilevel context. In fact, Kaufmann& Sydney(2020) argued that MLG should be part of any study of urban policy. This is because municipal policymakers act at the intersection of vertical and horizontal axes. First, city officials’ scope for formulating policy responses is shaped by formal and informal institutional constraints. Second, local non-state actors – such as platform workers, unions, or platform managers – might demand that their views be reflected in municipal policies. Yet, acquiring a comprehensive picture of how city governments steer platform work requires an appreciation of municipal officials’ perception of the issue, or the motivation for engagement in respective debates. That perception could be influenced by the vertical axis(if another governmental level is understood to address challenges related to platform work) or horizontal axis(if an interest group successfully promotes its framing of platform work). Previous studies, for instance by Ardura Urquiaga et al.(2019), have pointed to the significance of municipal perceptions, though it remains to be investigated how these interact with the interests of nonstate actors. In other words, empirical evidence suggests that city governments do not merely exercise the competences granted to them and automatically react to the framing of platform work by other urban stakeholders. Instead, they are more autonomous in choosing their reaction to app-based work. municipal, regional, national – regarding which one is responsible for tackling the issue. This relationship need not be formalised: there could be tacit agreement, or an expectation by the municipal government that the state, through appropriate legislation, will resolve the tensions surrounding platform workers. Third, focusing on the availability of competences and instruments at municipal level should tell us something about the formal constraints that the city government needs to navigate, as well as whether it circumvents the lack of competences pertaining to platform work through“soft”(Gupta et al., 2015: 140) or atypical policy tools, for example the use of incentives, negotiation, or mediation. Fourth, observing the involvement of the city government and its relevant officials with regard to this issue should consider the reasons why they choose(not) to engage. This indicator hence sheds light on a municipality’s policy priorities and whether platform work is considered to be a concern. Such focus can expose the more ideological foundations of why a city administration is more engaged in the subject than other actors, depending on which role city officials think appbased work should play in the local economy. The following analysis of empirical data through MLG is based on 17 semi-structured interviews with state and nonstate actors as well as academic experts. Interviews served to illuminate the roles, views, aims, and tools of governmental entities(municipal, regional, federal, or state governments) and of non-state actors(platform workers’ representatives and unions) in the governance of platform work in Madrid, Milan and San Francisco. Additionally, documentary analysis of municipal policy papers was used to perform data triangulation. All data was subjected to inductive as well as deductive rounds of qualitative coding. The final step in mobilising MLG in order to apply it to municipal policy consists of its operationalisation through indicators that can be observed(Seeber, 2020) and helping “contextualise empirical observations”(Martinez in Allen, 2017). Based on the preceding discussion, these indicators are: 1. roles, views, and involvement of non-state actors; 2. the relationship between governmental layers regarding platform work; 3. the availability of competences and instruments at the municipal level; and 4. the involvement of the municipality in the policy issue of platform work. First, it needs to be studied which non-state actors seek to be involved in respective policy processes, which tools they employ, which governmental level they try to engage with, and to what effect. A crucial aspect of these actors is their interaction with governmental authorities at other levels, and the reasons for such. Second, observing the relationship between governmental layers regarding app-based work should indicate where the main regulatory response to the phenomenon takes place, and whether there is some agreement between the layers – 3.4 RESULTS THREE DISTINCT MULTI-LEVEL GOVERNANCE CONTEXTS By applying the four MLG indicators to qualitative data collected in Madrid, Milan, and San Francisco, it is analysed how the interplay of these indicators explains respective municipal response to platform work. MADRID(SPAIN) The policy outcome – the unwillingness of Madrid’s city government(Ayuntamiento de Madrid) to recommend platform work to local jobseekers – can only be explained through the combination of all four MLG indicators. Against the backdrop of Spain’s nationional Rider Law, initiated by the country’s Labour Ministry in 2021 to classify platform workers as employees, concerns among city officials about steering jobseekers into precarious forms of work and the self-perception as pilot city for the Rider Law produce this specific municipal response. In the context of constrained municipal competences, limited capacities to manage tensions between workers and platforms and focus on other 18

PLATFORM WORK IN MADRID, MILAN, AND SAN FRANCISCO local challenges – above all high levels of underemployment and unemployment in the city – the decision not to recommend app-based work constitutes a simple, yet significant, step. It furthers the municipality’s goal of combatting widespread precarity in Madrid’s labour market without the need to cooperate with other stakeholders or gain their consent, in contrast to Milan. The perception of platform work as not being an urgent issue also supports an earlier finding (Ardura Urquiaga et al., 2019) that municipal officials do not perceive the platform economy to be a major concern. In the face of the hierarchical governance structure that concentrates actions on platform work at the national level, namely through the Rider Law, non-state actors’ efforts to influence the conditions underlying app-based work in Madrid have yielded only limited success. As a result of both necessity and intent, these efforts are confined to two other courses of action. First, they resort to engagement with the regional government, which holds no formal competences pertaining to platform work, but has been open to dialogue with stakeholders. Second, non-state actors make use of alternative channels. These include the organisation of protests as well as the creation of collectives and cooperatives in the case of workers, and the formation of interest groups, such as Adigital and APS, that seek to shape public opinion in the case of platforms. Therefore, Madrid’s case study is particularly insightful considering the power constellation it reflects, with state and municipal governments being able to respond to the precarious conditions of platform workers by limiting the influence of platforms and respective interest groups in governance processes. MILAN(ITALY) The municipal management(Comune di Milano) of appbased work here consists of a proactive and interactive approach that engages with relevant stakeholders and seeks local solutions. The response results from the general municipal tendency to adopt interactive governance modes that rely on non-state actors to implement policies, as well as from City Hall’s commitment to the protection of workers’ rights, a legacy of Milan’s industrial roots. The Milanese response to platform work also constitutes a reaction to public initiatives at other governmental levels. On one hand, the insufficiency of national attempts to reconcile workers and platforms through consultations instigated by the Italian Labour Ministry has pushed the municipality to respond to the public protests of app-based workers and initiate its own monthly roundtable with social partners and platform companies. This constitutes an attempt to ensure both the continuation of innovative platform services and the protection of workers. In fact, the roundtable has resulted in changes in municipal transport rules following workers’ requests. On the other hand, Milan’s city government finds itself in competition with other Italian municipalities, including Naples and Torino, over which city manages platform work most effectively. These two cities were the first to institute so-called Rider Spots, where workers receive support and guidance from city officials, prompting the municipality of Milan to propose a similar concept. MLG also shows that workers and their representatives, and platform companies and their Italian interest group AssoDelivery, do not confine themselves to the city level. The former pursue several avenues to promote their views, including protests, court cases, and mobilisation of platform workers, while the latter focus their attention at the national level to affect potential relevant legislation. All in all, non-state actors play significant roles in Milanese governance processes involving platform work thanks to the national and municipal roundtables – which offer them portals into responsible policymaking and legislative chambers – but also due to the nature of the proposed Rider Spot, relying on platforms to finance the project. SAN FRANCISCO(USA) The municipality of San Francisco(City and County of San Francisco) offers a favourable environment for experimentation with, and use of, platforms. Yet high costs of living and mounting inequality have caused poverty and homelessness on an unprecedented scale. Actions of the city government attempt to balance these tensions, on one hand attracting platform companies while upskilling jobseekers on the other. Within this context, platform work is seen as opportunity and recognised as a positive outcome for jobseekers looking to earn a living by San Francisco’s municipal department for labour policy, OEWD. As such, platforms perform an important function in the municipal labour strategy. Still, City Hall is not a passive bystander in the platform economy: activities of platforms are subject to municipal regulation, though not in terms of work, where city competences are restricted. For example, the municipality subjects ride-hailing platforms to a congestion tax. Platforms have so far prevailed in debates concerning the governance of app-based work in San Francisco thanks to a unique mobilisation of financial resources and political lobbying at various governmental levels, which is also confirmed by earlier studies(McNeill, 2016; Walker, 2018). Assembly Bill 5(AB 5), aiming to classify app-based workers as employees and thus grant them access to employment-related social benefits, epitomises both the attempts of California’s state government to address the precarity of workers, but also some platforms’ success in being exempted. However, ongoing tensions between policymakers, platforms and workers suggest that platforms’ dominance in deciding the future of San Francisco’s app-based workers remains contested. For example, state-level lawmakers are still seeking to enforce remaining parts of AB 5. GOING ABOVE AND BEYOND THE AVAILABILITY OF COMPETENCES Compared to earlier studies on the governance of platform work in cities, the application of MLG points to the significant role of the perception of app-based work by municipal officials and their subsequent willingness or reluctance to engage with the tensions arising from its growth. Such perception can best be explained in tandem with other indicators, however. The case of Madrid suggests that its municipal officials do not consider platform work to be an urgent policy concern due to different policy priorities, including underemployment and unemployment in the local labour market. Yet, their lack of concern likely also reflects officials’ awareness of the national Rider Law aiming at correcting the misclassification of platform workers, thus taking pressure to intervene off the city government. In contrast, the munic19

FRIEDRICH-EBERT-STIFTUNG – TRAJECTORIES OF PLATFORM CAPITALISM AND PLATFORM WORK ipality of Milan operates within a highly diverse governance environment, where the lack of a comprehensive national response to platform work and a municipal desire to balance innovation and workers’ protection push the city administration to engage with non-state actors in search of local solutions. Meanwhile, not only do San Francisco’s municipal officials consider platform work to be a steppingstone for jobseekers – they regard it as remedy for residents to make ends meet in view of an exorbitant cost of living. These findings in turn put in question the assumption that granting city governments more extensive competences on labour issues and platform work would necessarily improve workers’ conditions and welfare entitlements. The factors influencing municipal responses to app-based work are more variegated than simply relating these to an outcome of greater or fewer competences. The foregoing analysis indicates that city governments hold crucial roles in municipal debates surrounding platform work, even if limited legislative competences or opposition by platforms suggest otherwise. This is because they can put in place non-legislative measures that address aspects of workers’ precarity, as Milan’s Rider Spot demonstrates, or facilitate a dialogue between workers and platforms. By the same token, they can contribute to the presence of platform work by recognising it as positive outcome for jobseekers, as in San Francisco. The findings also indicate that platform governance and platform work governance in cities are not necessarily the same. San Francisco shows how the municipality is willing to impose a congestion tax on ride-hailing platforms, while welcoming work on the same platforms. This, again, can best be explained through an analysis of the four MLG indicators, and by exploring how the combination of limited competences, perception of appbased work by municipal officials and platforms’ lobbying efforts combine to influence the policy outcome. 3.5 CONCLUSION Multi-level governance(MLG) offers researchers of platform urbanism a key to understanding those complex elements – state and non-state actors’ influence, competences, governmental structures – that influence the scope and nature of municipal responses to platform work. Crucially, the cases of Madrid, Milan, and San Francisco emphasise the interplay between these elements, and how they can reinforce each other. The framework now needs to be tested and modified by applying it to other cities. Further research could moreover evaluate MLG’s normative potential. By employing indicators to reveal which factors hinder greater protection of workers at the urban level, MLG could point to the direction where change needs to happen – an urgent task in the face of platform workers’ continuing precarity. 20

PLATFORM WORK IN MADRID, MILAN, AND SAN FRANCISCO BIBLIOGRAPHY Aguilera, T., Artioli F.& Colomb, C.(2019). Explaining the diversity of policy responses to platform-mediated short-term rentals in European cities: A comparison of Barcelona, Paris and Milan. EPA: Economy and Space, 53(7), 1689–1712. Allen, M.(ed.)(2017). The SAGE Encyclopedia of Communication Research Methods. Thousand Oaks: Sage Publications. Ardura Urquiaga, A., Lorente-Riverola, I., Mohino, I.& Ruiz Sanchez, J.(2019).‘No estamos tan mal como Barcelona’: análisis de la proliferación y regulación de las viviendas de uso turístico en Madrid y Barcelona. Boletín de la Asociación de Geógrafos Españoles, 83(2828), 1–47. Bache, I.& Flinders, M.(eds.)(2004). Multi-Level Governance. Oxford: Oxford University Press. Barns, S.(2014). Platform urbanism: The emerging politics of open data for urban management. Paper presented at the Association of American Geographers Annual Meeting, Tampa, FL, 08–12 April. Beer, R., Brakewood, C., Rahman, S.& Viscardi, J.(2017). Qualitative Analysis of Ride-Hailing Regulations in Major American Cities. Transportation Research Record: Journal of the Transportation Research Board, 2650, 84–91. Bessa, I., Joyce, S., Neumann, D., Stuart, M., Trappmann, V.& Umney, C.(2022). A global analysis of worker protest in digital labour platforms, ILO Working Paper 70. Geneva: ILO. Bond, A.(2015). An App for That: Local Governments and the Rise of the Sharing Economy. Notre Dame Law Review, 90(2), 77–96. Brail, S.(2018). From Renegade to Regulated: The Digital Platform Economy, Ride-hailing and the Case of Toronto. Canadian Journal of Urban Research, 27(2), 51–64. Bramwell, A.(2020). Innovation and the“Ordinary” City? Urban Policy Making in a Digital Age. American Political Science Association Politics Symposium, 15–19. Caprotti, F., Chang, I.& Joss, S.(2022). Beyond the smart city: a typology of platform urbanism. Urban Transformations, 4(4). Cherry, M.& Aloisi, A.(2017).‘Dependent Contractors’ in the Gig Economy: A Comparative Approach. American University Law Review, 66(3), 635–690. Da Cruz, N., Rode, P.& McQuarrie, M.(2019). New urban governance: A review of current themes and future priorities. Journal of Urban Affairs, 41(1), 1–19. Flores, O.& Rayle, L.(2017). How cities use regulation for innovation: the case of Uber, Lyft and Sidecar in San Francisco. Transportation Research Procedia, 25, 3756–3768. Gupta, J., Pfeffer, K., Verrest, H.& Ros-Tonen, M.(eds.)(2015). Geographies of Urban Governance Advanced Theories, Methods and Practices. Springer International Publishing Switzerland. Hooker, J.& Antonucci, L.(2022). Improving the EU Platform Work Directive proposal: a contribution from emerging research findings, OSE Paper Series, Opinion Paper No. 28. Brussels: European Social Observatory. Kaufmann, D.& Sydney, M.(2020). Toward an Urban Policy Analysis. American Political Science Association Politics Symposium, 1–5. Marks, G.(1993). Structural Policy and Multilevel Governance in the EC. In Cafruny, A.& Rosenthal, G.(eds.), The State of the European Community. Boulder: Lynne Rienner. McNeill, D.(2016). Governing a City of Unicorns: Technology Capital and the Urban Politics of San Francisco. Urban Geography, 37(4), 494–513. Morell, M.(ed.)(2018). Sharing Cities: A worldwide cities overview on platform economy policies with a focus on Barcelona. Barcelona: Universitat Oberta de Catalunya. Ongaro, E.(ed.)(2015). Multi-Level Governance: The Missing Linkages. Emerald Group Publishing Limited. Prassl, J.(2018). Humans as a Service: The Promise and Perils of Work in the Gig Economy. Oxford: Oxford University Press. Sadowski, J.(2021). Who owns the future city? Phases of technological urbanism and shifts in sovereignty. Urban Studies, 58(8), 1732–1744. Seeber, M.(2020). Framework and operationalisation challenges for quantitative comparative research in higher education. Higher Education Quarterly, 74, 162–175. Tabascio, A.& Brail, S.(2021). Governance matters: Regulating ride hailing platforms in Canada’s largest city‐regions. The Canadian Geographer/Le Géographe canadien, 1–15. Thelen, K.(2018). Regulating Uber: The Politics of the Platform Economy in Europe and the United States. Perspectives on Politics, 16(4), 938–953. Vith, S., Oberg, A., Höllerer, M.& Meyer, R.(2019). Envisioning the ‘Sharing City’: Governance Strategies for the Sharing Economy. Journal of Business Ethics, 159, 1023–1046. Voytenko Palgan, Y., Mont, O.& Sulkakoski, S.(2021). Governing the sharing economy: Towards a comprehensive analytical framework of municipal governance. Cities, 108. Walker, R.(2018). Pictures of a Gone City: Tech and the Dark Side of Prosperity in the San Francisco Bay Area. PM Press. Wolf, A.(2022). City Power in the Age of Silicon Valley: Evaluating Municipal Regulatory Response to the Entry of Uber to the American City. City& Community, 00(0), 1–24. Zanatta, R.& Kira, B.(2018). Regulation of Uber in São Paulo: from conflict to regulatory experimentation. International Journal of Private Law, 9(1). 21

IMPRINT ABOUT THE AUTHORS IMPRINT Milena Franke is a doctoral researcher and teaching assistant at KU Leuven, Belgium. Her doctoral research project focuses on digital labour in Belgium and seeks to understand worker control, consent and unpaid work in the platform economy. MIlena works on the ERC-AdG project ‘ResPecTMe’, the FWO project‘Precarious work in the online economy: A study on digital workers in Belgium and the Netherlands’ and the‘Fairwork’ project, which is based at the Oxford Internet Institute and the WZB Berlin Social Science Center. Valentin Niebler is a doctoral researcher at the Institute for European Ethnology(IfEE) and at the Berlin Institute for Migration Research(BIM) at Humboldt University of Berlin. His dissertation project investigates the formation of white-collar worker organisation at tech companies. Previously, Valentin has worked as research associate on the project Platform Labor in Urban Spaces(PLUS) at Humboldt University of Berlin. Maximilian Kriz received a PhD degree in Urban Studies from the University of Glasgow in 2023. His doctoral thesis explored platform work in municipal contexts in Madrid, Milan, and San Francisco. Maximilian continues to work at the intersection of policymaking and development at the intergovernmental level. Published by Friedrich-Ebert-Stiftung| Competence Centre on the Future of Work| Cours Saint Michel 30a| 1040 Brussels| Belgium Matthias Weber, Head of Friedrich-Ebert-Stiftung Competence Centre on the Future of Work Responsible Contact: Inga Sabanova inga.sabanova@fes.de For more information about the Competence Centre on the Future of Work, please consult: https://futureofwork.fes.de/ Design: Petra Strauch/Barbora Novotna The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung. Commercial use of media published by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES. Commercial use of all media published by the Friedrich EbertStiftung(FES) is not permitted without the written consent of the FES ISBN: 978-3-98628-345-2 © 2023 www.fes.de/bibliothek/fes-publikationen