Pressure and praise as an action research methodology: The case of Fairwork

Graham, Mark et al. Article — Published Version Pressure and praise as an action research methodology: The case of Fairwork Environment and Planning A: Economy and Space Provided in Cooperation with: WZB Berlin Social Science Center Suggested Citation: Graham, Mark et al. (2025) : Pressure and praise as an action research methodology: The case of Fairwork, Environment and Planning A: Economy and Space, ISSN 1472-3409, SAGE Publications, London, Iss. OnlineFirst, pp. 1-15, https://doi.org/10.1177/0308518X251336893 This Version is available at: https://hdl.handle.net/10419/325856 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/4.0/ https://doi.org/10.1177/0308518X251336893 EPA: Economy and Space 1 –15 © The Author(s) 2025 Article reuse guidelines: sagepub.com/journals-permissions DOI: 10.1177/0308518X251336893 journals.sagepub.com/home/epn AEconomy and Space Pressure and praise as an action research methodology: The case of Fairwork Mark Graham Oxford Internet Institute, University of Oxford, Oxford, UK Oğuz Alyanak Oxford Internet Institute, University of Oxford, Oxford, UK Alessio Bertolini Oxford Internet Institute, University of Oxford, Oxford, UK Patrick Feuerstein Research Group Globalization, Work and Production, WZB Berlin Social Science Center, Germany Tobias Kuttler Research Group Globalization, Work and Production, WZB Berlin Social Science Center, Berlin, Germany Funda Ustek Spilda Department of Digital Humanities, King’s College London, London, UK Jonas Valente Oxford Internet Institute, University of Oxford, Oxford, UK Abstract In platform-mediated sectors such as ride-hailing, delivery, care work and cloudwork, gig workers often lack essential employment protections such as minimum wage and social security. Misclassification of workers, opaque algorithms, exploitation of legal loopholes and lobbying by platforms against protective legislation further exacerbates inequalities and discrimination. The Fairwork project at the University of Oxford and WZB Berlin, in collaboration with a global network of partners, aims to improve conditions in the platform economy through action research. Using a principles-based framework, the project evaluates and scores platforms against universal standards of fairness – fair pay, fair conditions, fair contracts, fair management and fair representation. The actionCorresponding author: Mark Graham, Oxford Internet Institute, University of Oxford, 1 St Giles, Oxford OX1 3JS UK. Email: [email protected] 1336893EPN0010.1177/0308518X251336893EPA: Economy and SpaceGraham et al. research-article2025 Exchanges 2 EPA: Economy and Space 00(0) research methodology combines data triangulation and stakeholder engagement to ensure platform evaluations are objective and impartial, while seeking to hold platforms accountable and advocating for better labour standards in the gig economy. This article explores Fairwork as an action research methodology that has established positive feedback loops, driving platforms towards pro-worker policy changes. We hope this discussion encourages other politically engaged social scientists to adopt action research approaches in their own work. Keywords Platform work, platform economy, digital labour, gig economy, action research, digital capitalism, research methods What is at stake for workers in the gig economy? There are millions of workers in the ‘gig economy’, who undertake short-term platform-mediated jobs like food delivery, driving, cleaning and care work.1 Proponents of the gig economy foreground the flexibility and autonomy that can be enjoyed by its workers; however, in many ways, platform work presents characteristics of traditional temporary and casualised work, such as low pay, instability and limited employment protection (Woodcock and Graham, 2020). As most platform workers are classified as self-employed, they tend to fall outside the scope of most employment rights and protections, including minimum wage legislation, working time regulation and health and safety protection (Fredman et al., 2021). The opaque algorithmic management of platform workers, with algorithms generally operating as a ‘blackbox’, with no transparency or accountability (Aneesh, 2009; Muldoon and Raekstad, 2023), leaves workers in a subordinate position when trying to make claims against platforms or the State (Kuttler, 2023). The use of opaque algorithms can also drive gender, racial and class discrimination based on customer ratings (Teng et al., 2023) and can compel low-income and minoritised workers to accept low wages (Dubal, 2023). Variable and changing bonus systems and gamification tools further increase the risks of overwork and excessive risk-taking on the job (Vasudevan and Chan, 2022). Platforms also harvest vast amounts of data from workers, often for commercial purposes, with limited external scrutiny or accountability (van Doorn and Badger, 2020). Finally, because of the atomised and isolated nature of platform work, workers often find it difficult to build collective power and to find a collective voice (Alyanak et al., 2023; Bertolini and Dukes, 2021). The establishment of formal and structured collective bargaining initiatives has been slow, and while worker organising has taken place, precious few victories have been won (Alyanak and Karlıdağ, 2022; Bonhomme et al., 2024; Howson et al., 2020) and the significant power imbalance between platforms and workers has hampered concerted interventions directed at these companies, aiming for better working conditions and payments (Lopez et al., 2024). Although there are established frameworks for decent work, most notably the ILO’s Decent Work Agenda, they do not address the specificities of platform work, especially in the absence of established legal frameworks. Many scholars have highlighted that, in most jurisdictions, platform regulation has been inadequate or ineffective (Bertolini, 2024; Collier et al., 2017; Makó et al., 2022; Stewart and Stanford, 2017), with a broad coalition of economic, political and academic stakeholders, as well as civil society organisations, noting the power imbalances in the relationship between platform firms and workers. Platforms have also invested staggering sums in lobbying for deregulation or regulatory change in their favour (Gromis, 2021; MOVE EU, 2021, also Naumov and Keith, 2023). The ‘Uber Files’ investigation conducted by a consortium of media outlets revealed the strategies employed by Uber to co-opt authorities and lobby to ensure the passing of platform-friendly regulations around the world (Corporate Europe Observatory, 2022). In the US, platforms spent over $200 million in their fight against the bill in California that aimed to reclassify platform workers as employees, undertaking what Shapiro (2023) referred to as ‘platform sabotage’. Similarly, in 2023, the Estonian-based ride-hailing platform Bolt lobbied the Estonian Graham et al. 3 government to reject the legal presumption of employment in the EU Directive on Improving Working Conditions in Platform Work (Bourgery-Gonse, 2024; EUObserver, 2024). Although a compromise version of the Directive was passed eventually in March 2024, it is thought that lobbying activity undertaken by platforms significantly weakened its content (Corporate Europe Observatory, 2024). The rise of action research methods in social sciences and economic geography Action researchers position themselves as instigators of change – within the academy, but also in their relations with research participants. They seek a new relationship between knowledge production and practice and are often very vocal about their aim of strengthening communities and collective action. Since action research builds on a range of liberal, emancipatory and critical traditions in the North and South, for action researchers to care for research participants and their communities is not in contradiction to good research practice and rigorous methodology, rather the opposite (Bradbury, 2015; Cornish et al., 2023; McNiff and Whitehead, 2013; Rowell et al., 2017). Action research is the result of disenchantment with the claims of positivist research to be objective and rational. The efforts of social scientists to stay neutral have resulted in the research being passive to questions of profound social change, and also in detachment from research participants and communities, leading to questioning of the relevance of social sciences to human concerns (Greenwood, 2002; Somekh, 2003). Despite these shortcomings of conventional social sciences, approaches such as action research that seek to be reflexive and socially accountable are seen as ‘risky’ or denounced as ‘activism’. This is not surprising, since positivist worldviews continue to dominate most social science research. Greenwood (2017: 181) notes that ‘Academic social scientists are trained in, and mainly obey, the hegemonic disciplinary structures in their universities. Many of those employed by or in the private and public sector generally view their work as fee-for-service analysis and ‘decision support’. In both cases, rocking the boat with democratically-inspired social critiques or actively promoting social reforms [. . .] is rare and is often punished’. While action research has tended to sit outside of the mainstream of objectivist, positivistic Westerncentric research, in recent decades it has made inroads into many disciplines, including development, geography, management, social work, sociology, anthropology, architecture, planning, healthcare and education. In the face of climate change and the destruction of ecosystems, but also the uncertainties that are associated with digitation and AI, it is becoming increasingly evident that current means of knowledge production and dissemination are insufficient and that translation of knowledge into practice is failing. Hence, action researchers have been constantly pushing the academic community to turn towards more action-oriented research based on experimental and collaborative learning, potentially opening up pathways for more desirable futures (Bradbury et al., 2019; Rowell and Call-Cummings, 2020). Economic geography has also experienced a shift of attention to the local scale and local actors. As Bjørnar Sæther noted in 2006, it has long been ‘hard to find researchers within economic geography who acknowledge that participants co-generate knowledge’. However, increased interest in understanding innovation and regional development has shifted the research focus to the local level (Dicken, 2003), and the study of networks and knowledge transfer in regional development has put a spotlight on the logics and perspectives of embedded actors (Grabher, 2004; Wittel, 2001). Based on such a broadening of scope, action research approaches have been particularly salient in geographic development research and adjacent disciplines that study rapid change in global value chains and the effects on local labour markets. Through action research, scholars have worked with local communities and grassroots organisations to expose worker exploitation and poor labour conditions (Benstead et al., 2018), while also illuminating worker resistance, organisation and empowerment (Fuentes, 2023). Action research methodologies make accessible the knowledge and perspectives of the involved or affected actors who would otherwise be overlooked in conventional research. Such 4 EPA: Economy and Space 00(0) knowledge was proven valuable in informing strategies to hold corporations accountable, foster ethical economic practices in supply chains, and advocate for stronger international labour protections (Hale and Wills, 2007). Dickson (2013: 134) aptly summarises the benefits of action research, while pointing to the dilemma of conventional approaches: ‘[A]cademic research is slow. [. . .] Undoubtedly, the traditional academic research is important [. . .] but we need quicker action. As the years have passed while studies carried out, workers have continued to be exploited and denied their rights.’ Action research in the gig economy: The Fairwork project The Fairwork project at the University of Oxford and WZB Berlin, in collaboration with a global network of partners, aims to improve conditions in the platform economy through action research. Since 2018, it has built a principles-based framework to describe and help address the risks and harms workers experience in the fast-growing platform economy, and the mechanisms that platforms should put in place to guarantee workers a minimum standard of fair work. The core idea of the Fairwork project is that companies can be encouraged to improve working conditions when they are scored against universal standards of fairness and ranked against each other. Because few platforms actually employ their workers or own much in the way of fixed capital, the idea is that they are especially vulnerable to public pressure (Graham, 2020). Consumers can relatively easily switch from one company to another. To deploy this sort of ranking and scoring strategy, however, requires definition of universal labour standards, that is, a set of thresholds which can be used to evaluate working conditions irrespective of the sector within which a company operates. The project methodology was designed in three stages: (1) develop a framework to evaluate platforms against principles of fairness; (2) evaluate platform companies against those principles and provide them with an easy to understand score; (3) use the resulting score of fair working conditions to compare companies against one another, and – in doing so – encourage companies to improve their scores, and thus the earnings and working conditions of workers. To achieve the first objective, in 2018 the project organised a workshop at the International Labour Organisation and United Nations Conference on Trade and Development in Geneva, which brought together key stakeholders from worker groups, unions, academia and platforms (Graham et al., 2020). In this workshop, five core principles of fair work were developed: fair pay, fair conditions, fair contracts, fair management, and fair representation. The principles were first refined through multistakeholder workshops in South Africa, India and Germany, and later in sessions held globally. These workshops prioritised input from unorganised workers, who represent the vast majority of platform labour forces. Five universal principles of Fair Work Two key considerations were foregrounded in the development of the five principles. First, they needed to be operationalised in evidence-based ways. Each of the five principles entails two thresholds of criteria that platforms can be evaluated against, with platforms being awarded an overall score out of 10 (see Figures 1–3). Second, the principles needed to be universal. One common argument from platforms in response to complaints about the quality of the jobs they offer is that they provide better work alternatives than exist in much of the Global South. It is also argued that platform work provides a pathway out of poverty – a claim that is hard to substantiate by empirical evidence (Anwar and Graham, 2021; Cook and Rani, 2023; Soriano and Cabañes, 2019). Although the authors agree that the conditions in the informal sector of, for instance, cleaners in South Africa and delivery workers in India are generally poor, we maintain that large corporations can, and should, do much more to protect the welfare and wellbeing of their workers. Graham et al. 5 The Fairwork project team has therefore operationalised the five principles in such a way that they apply to all platform workers, in all sectors and all places, under all types of contracts. This was no easy task.2 Following the stakeholder workshop at the ILO, the principles were taken to in-country workshops in India, South Africa and Germany in 2018 and 2019. Workers, unions, and platform managers were asked to discuss their priorities for the principles and rate their importance. The research team then synthesised the findings into robust minimum labour standards. As the project expanded in subsequent years to working in 39 countries by 2024, that task of listening to diverse stakeholders worldwide – including unions, worker groups, regulators, platform managers and consumer groups – has grown ever more challenging. Regular and extensive engagement with these stakeholders led to subsequent rounds of systematic refinement of the principles. The final set of principles that is applied in Fairwork research (Fairwork, 2025a) reflects these efforts to set universal standards, grounded in very specific cultural-, classed-, gendered-, placeand economic sector-specific concerns about job quality and decent work3 (see Figures 1–3). Putting the Fairwork principles into practice The Fairwork method evaluates and publicly ranks platform companies, and as such, it must be objective and impartial. Points are only awarded to platforms when the research team finds reliable evidence that the threshold has been met. That means that a point not awarded to a platform does not necessarily mean that the threshold is not met, but rather that we could not find sufficient evidence that it was (Graham and Anwar, 2018). Figure 1. Fairwork principles. Source. Fairwork (2025a). 6 EPA: Economy and Space 00(0) Figure 2. The two principle thresholds for Principle 4 on Fair Management (for Location-based platform work). Source. Fairwork (2025b). To assess whether a threshold is met, the team triangulates data from multiple sources (Fairwork, 2025c). The research data is collected through desk research (reviewing all information publicly available on the platform, including contracts and terms and conditions, insurance policies if they exist, public statements, etc.), interviews or online surveys with platform workers, and interviews with platform managers. This approach allows us to integrate different perspectives and crosscheck the information received from each source to develop a thorough assessment of the working conditions on that platform. As such, rather than a simple audit to verify elements of company practice and compliance, Fairwork creates a large repository of rigorous academic research on the platform economy.4 Needless to say, academic rigour is indispensable in triangulating the evidence and ensuring that all elements of research follow the strictest academic ethics guidance and rules. Fairwork research has received ethical clearance from the University of Oxford’s Departmental Research Ethics Committee (DREC) and seeks ethical clearance in all the locations where it operates. A series of mandatory processes and checks has been built into the methodology to validate the results and prevent subjective bias. Within the research team, ‘country liaisons’ are responsible for consistency of the operationalisation process within a country, and ‘platform liaisons’ are responsible for the operationalisation process for platforms that operate across multiple countries. These roles were not envisioned in the early stages of the project, but as the project expanded to 41 countries (with Graham et al. 7 Figure 3. Fairwork scoring system. Source. Fairwork (2025a). over 710 company scorings conducted at the time of writing), attention to consistency of scoring became increasingly essential in view of the tremendous variety of business models and contexts in which platforms operate. Internal peer review by members of the Fairwork network further maximises consistency of platform scorings and the yearly country reports. As part of the peer-review process, learnings are recorded in an ‘operationalisation’ document that highlights cases in the scoring process that do not neatly fulfil the principle thresholds – for example, due to unexpected clauses in policies or ways of practicing them. This ensures that precedents can be followed in future scoring rounds. Given the wide variety of platform models, practices and policies and their different application in different jurisdictions, the operationalisation documentation guides the team in ensuring validity and reliability across scoring decisions. Turning research into action How, then, can those scores be leveraged to improve pay and conditions for workers in the platform economy? At the onset of each annual scoring cycle, the Fairwork country teams host workshops with stakeholders, including platform managers, unions and worker organisations, to present the previous platform scorings, the Fairwork methodology and the research timeframe and steps (Alyanak et al., in press). These events are not just about information sharing but also about gathering valuable feedback from platforms, unions and other relevant stakeholders. During the research process, platforms are repeatedly consulted and encouraged to make pro-worker changes to their processes and policies. Before publishing the final scorecards, the provisional scores are shared with the platform management, with detailed explanations about how the policies and practices can be improved to meet the minimum standards of decent work as described in the Fairwork principles. They are given 6–8 weeks to make additional changes to their policies before the final platform ratings are published. 8 EPA: Economy and Space 00(0) The final ratings make the companies at the top and the bottom of each country’s score table easily identifiable. The Fairwork team publishes press releases and contacts media representatives and policy makers in order to ensure maximum dissemination. As a result, platforms with low scores are encouraged to positively engage with the research team in future scoring rounds. To give one example of this, in India in the first year of research in 2019 only 2 out of 12 companies engaged with the research team. By the second year, this had risen to 4 out of 11. In the most recent year, 2024, 7 out of 11 companies engaged with us (Figure 4). This positive feedback loop centred around platforms’ reputation has led to significant improvements in platforms’ policies and practices, reflected in their improved ratings. At the time of writing in early 2025, over 320 pro-worker policy changes have been made by the platforms as a result of their engagement with the Fairwork team. Given that some platforms host hundreds of thousands of Figure 4. X (formerly Twitter) thread written by Deepinder Goyal, CEO of Indian platform Zomato. Source. Goyal (2020). Graham et al. 15 and Informality: Pathways of Change, Alteration, and Transformation. Cham: Springer International Publishing, pp.89–138. Littler J (2013) What's wrong with ethical consumption? In: Lewis T and Potter E (eds) Ethical Consumption: A Critical Introduction. London: Routledge, pp.27–37. Lopez T, Jesnes K, Alyanak O, et al. (2024) Building labour power in the platform economy: A comparative analysis of worker struggles in German and Norwegian food and grocery delivery. New Technology Work and Employment. Epub ahead of print 25 October 2024. DOI: 10.1111/ntwe.12312 Makó C, Illéssy M, Pap J, et al. (2022) Emerging platform work in the context of the regulatory loophole (The Uber fiasco in Hungary). Journal of Labor and Society 26(4): 533–554. Martin-Caballero A (2024) New regulations on platform work: Fragmented responses to issues of work fragmentation. Capital & Class. Epub ahead of print 27 April 2024. DOI: 10.1177/03098168241248990. McNiff J and Whitehead J (2013) Action Research: Principles and Practice. London: Routledge. MOVE EU (2021) Our positions. Available at: https://www.move-eu.eu/our-positions (accessed 19 June 2024). Muldoon J, Graham M and Cant C (2024) Feeding the Machine: The Hidden Human Labour Powering AI. London: Cannongate. Muldoon J and Raekstad P (2023) Algorithmic domination in the gig economy. European Journal of Political Theory 22(4): 587–607. Naumov S and Keith D (2023) Optimizing the economic and environmental benefits of ride-hailing and pooling. Production and Operations Management 32(3): 904–929. Niebler V, Pirina G, Secchi M, et al. (2023) Towards ‘bogus employment?’ The contradictory outcomes of ridehailing regulation in Berlin, Lisbon and Paris. Cambridge Journal of Regions Economy and Society 16(2): 289–301. Rainone S and Aloisi A (2024) The EU Platform Work Directive: What’s New, What’s Missing, What’s Next?. ETUI Research Paper-Policy Brief. Available at: https://papers.ssrn.com/sol3/Delivery.cfm?abstractid=4957467 (accessed 18 March 2024). Reiser DB (2008) For-profit philanthropy. Fordham Law Review 77: 2437. Rowell L and Call-Cummings M (2020) Knowledge democracy, action research, the Internet and the epistemic crisis. Journal of Futures Studies 24(4): 73–82. Rowell LL, Bruce CD, Shosh JM, et al. (Eds)(2017) The Palgrave International Handbook of Action Research. New York: Palgrave Macmillan. Salvagni J, Grohmann R and Matos (2022) Gendering platform co-operativism: The rise of women-owned rider co-operatives in Brazil and Spain. Gender and Development 30(3): 707–724. Scheele L, Im ZJ and Leschke J (2023) Unpredictable and non-transparent working conditions?: riders in the food-delivery sector in six EU countries. EuSocialCit Working Paper, March 2023. Doi: 10.5281/ zenodo.7639909. Available at: https://helda.helsinki.fi/server/api/core/bitstreams/63bf6ee9-51e8-4ed386fc-9b0ad36905b2/content (accessed 17 March 2025). Scholz T (2016) Platform Cooperativism. Challenging the Corporate Sharing Economy. New York, NY: Rosa Luxemburg Foundation, p.435. Shapiro A (2023) Platform sabotage. Journal of Cultural Economy 16(2): 203–220. Somekh B (2003) Theory and passion in action research. Educational Action Research 11(2): 247–264. Soriano CR and Cabañes JV (2019) Between “world class work” and “proletarianized labor”: Digital labor imaginaries in the global south. In: Polson E, Schofield L and Gajjala R (eds) The Routledge Companion to Media and Class. London: Routledge, pp.213–226. Stewart A and Stanford J (2017) Regulating work in the gig economy: What are the options? Economic and Labour Relations Review 28(3): 420–437. Teng F, Botelho T and Sudhir K (2023) Can Customer Ratings be Discrimination Amplifiers? Evidence From a Gig Economy Platform. Yale University Working Paper, New Haven, Connecticut. van Doorn N and Badger A (2020) Platform capitalism’s hidden abode: Producing data assets in the gig economy. Antipode 52(5): 1475–1495. Vasudevan K and Chan NK (2022) Gamification and work games: Examining consent and resistance among Uber drivers. New Media & Society 24(4): 866–886. Wittel A (2001) Toward a network sociality. Theory Culture & Society 18(6): 51–76. Woodcock J and Graham M (2020) The Gig Economy. Cambridge: Polity.