Abstract
This article explores and develops the concept of ‘actually existing platformization’ through a case study of home-based care in Aotearoa New Zealand. We broaden its focus into a typology of four key sources of agency in the platformization process: policy, platform, workers and spatiality. We base our findings on two quite different care platforms: Mycare and My Homecare. Through document analysis, as well as interviews with care workers, we demonstrate how the four agencies shaped the two platforms in distinct and sometimes messy directions. Whilst the platformization process contributed to the rolling back of care worker rights and protections, whilst presenting frustrating barriers and even severe health risks, we also found it presented opportunities for the shared articulation of resistive care ethics, where a workforce comprized predominantly of women worked collectively to shape the path of platformization in their favour.
This article explores and develops the concept of ‘actually existing platformization’ (AEP) through a case study of home-based care in Aotearoa New Zealand (hereafter Aotearoa). Building on van Doorn et al.'s (2021) earlier conceptualization, we reiterate their argument that ‘things often do not go as planned’ (p. 715), and that scholars should consider the impact of local policy factors that can create requirements for ‘partnership building’ (p. 716) with established business and government interests. However, we argue that attention should also be afforded to the agency of workers and spatial factors in shaping the path of platformization. In the context of home-based care in Aotearoa, we found that workers affected platformization in two different ways: at the formal level through union representation, but also at the informal level of an ethics of care for their clients, which coalesced into networks of resistant solidarity. In terms of spatial factors, we found that Aotearoa's mountainous terrain and sparse population in rural areas both limited and shaped platformization in particular ways. These findings therefore add to a platformization of work literature which has so far predominantly focused on densely populated urban contexts. We formalize this broadened focus into a typology of four different sources of agency in the platformization process: policy, platform, workers and spatiality. Increased scholarly attention to broad sources of agency, which include both the human and the non-human, with some being less obvious or visible than others, we argue will be useful for future research on the platformization of work.
We base our findings on two quite different care platforms that have recently established themselves in Aotearoa: Mycare and My Homecare. The architecture of Mycare is more closely aligned with the features scholars have associated with care work platforms in other nations (Tandon and Rathi, 2021; Ticona and Mateescu, 2018), but its spread into the mainstream has been severely limited by Aotearoa's policy settings and worker organizing, forcing the organization to seek powerful partnerships in anticipation of a future funding regime change. My Homecare, by contrast, has developed a platform (along with its app for care workers, Geneva LIVE) that has been shaped by policy settings from the outset, incorporating some of the architecture of care work platforms, but without key agentive features.
Supplementing this analysis, we draw on 16 interviews and a focus group with Aotearoa-based care workers who use the platforms. They highlight both how platformization has impacted their working conditions but also how they are actively shaping the ways the technology is used and the meanings surrounding it through drawing on a feminist ethics of care. Further, they describe how spatial factors contributed to app dysfunction, with limited cell and internet network coverage creating interruptions and disconnections which influence the trajectory of platformization. However, firstly we define what we mean by both human and non-human agency in the platformization of work, in order to give more depth to our typology.
Conceptualizing agency in the platformization of work
With the new materialist or socio-materialist turn in social theory, scholarly attention has become increasingly attuned to the role of non-human actors in social life. Van Dijck's (2013) seminal book The Culture of Connectivity combines Critical Political Economy (influenced by Castell's theory of networks) with Actor Network Theory (influenced predominantly by Bruno Latour) to explore the spread of social media platforms into everyday life, facilitating new forms of social connection whilst also exploiting user connectivity (in the technical sense) as a resource for profit. Van Dijck's approach enables close attention to be paid to the materiality of technology and the agency of technology users, whilst simultaneously recognizing the role of political, economic and legal structures in shaping how particular configurations of human and non-humans take shape.
The new materialist and socio-material approaches disrupt the notion that agency, the capacity to act, is solely a human property. Agency, as conceived of in Western intellectual traditions, was theorized as a human property following Kant's proposition that they alone have the capacity for conscious, rational action. However, the new approach means conceptualizing agency as distributed: arizing from the complex and layered interactions between a host of human and non-human actors (Hjorth and Lupton, 2021; Lupton, 2017; Mos, 2021). Hence, for understanding the AEP of care work in Aotearoa, we have functionally delineated four sites of analysis which all combine, layer and distribute human and non-human agency: policy, platform, workers and spatiality. We hope this framework provides a basis for a better understanding the complex paths that platformization takes in given contexts.
Policy settings as agentive: the neoliberalized conditions of care provision in Aotearoa
In Aotearoa, since the neoliberal health reforms of the 1990s (see McGregor, 2001), home support for older people, as well as those with chronic conditions or disabilities, has been funded by the state but delivered largely by private organizations. The resulting administrative environment that delivers care is a complex bureaucratic public/private hybrid, resulting in significant variance across health conditions and regions, with highly distributed (rather than centralized) sources of agency. A variety of government bodies, such as the Ministry for Health, Te Whatu Ora and Te Aka Whai Ora (New Zealand's health authorities), Whaikaha (Ministry for Disabled People) and the Accident Compensation Corporation, contract an even wider array of private providers to deliver the care. These providers are a mix of for-profit and non-profit entities that compete against each other for contracts to deliver home-based care for populations in particular regions, through individualized per-client ratios. Pressured to provide the best service for a fixed price, providers are incentivized to drive down costs (Hennessy and Rodrigues, 2019), resulting in poor conditions for workers (Douglas and Ravenswood, 2019; Ravenswood et al., 2021). Further, an individualized funding system embeds underpayment for care workers, as per-client funding is passed from government departments to private providers, translating into piecemeal remuneration (despite holding permanent contracts with sick pay and annual leave). Shortfalls in funding are absorbed by clients whose needs are not met and by care workers who are not fairly compensated. The (mostly) women care workers therefore effectively subsidize ‘corporate profits and the taxpayer through their labour’ (Douglas and Ravenswood, 2019: p. 187).
Despite mirroring the gig economy in its low levels of pay, tied to a piecemeal remuneration system, in other ways, the majority of care work in Aotearoa has proved relatively insulated against the effects of platformization. Aotearoa's policy settings have ensured a relatively formalized sector (compared to other Organization for Economic Co-operation and Development nations) where a majority of care workers are classed as employees with permanent contracts, rather than self-employed contractors. This is more by accident than design – until 2022, Aotearoa's care sector was primarily funded through 20 District Health Boards (Tenbensel et al., 2017). In one sense, these boards were emblematic neoliberalized entities, in that they were classed as ‘independent’ from the central government whilst expected to balance their books, despite chronic underfunding (Barnett and Bagshaw, 2020). In another sense, however, they prevented public funding of home-based care from being placed into the hands of an individualized consumer who directly hires care workers as self-employed contractors. Rather, the District Health Boards (DHBs) contracted with a range of health service providers, which, despite often being private, for-profit companies, were required to hire care workers as employees, with associated rights and protections, including the option of contracting under one of the collective agreements negotiated by unions.
This means that ‘Uber-for-care’ style platforms are not yet widespread in Aotearoa, because the majority of the population who require home-based care (e.g. the elderly) cannot directly use the funding received from the government for their services. As we explore in the below case study of two care platforms, Mycare and My Homecare, this has shaped their form; with the former putting the onus on pursuing partnerships with government agencies to develop alternative income streams, whilst the latter can only mimic some of the architecture of gig economy platforms, but without key agentive features.
The agency of platform operators
The agency of platform operators is, of course, partly technological, and therefore non-human. The rise of the platform model was built on the emergence of Web 2.0 technology, together with the smartphone. This technological architecture made possible the centralization of the web and the concentration of economic and political power within a small number of predominantly US-based corporations (Benvegnù and Kampouri, 2021; Helmond, 2015; Srnicek, 2017). However, technology was not working alone; Pohle and Voelsen (2022: p. 15) use the term ‘techno-political configuration’ to note how specific human actors sought ‘to fundamentally reconfigure the dominant network logic of the Internet’, subordinating it ‘to a more centralized logic’.
Furthering this concentration of power, the technological architecture of the platform enables ‘walled gardens’ (Plantin et al., 2018) – closed ecosystems which allow tight control of user access together with the smooth collection of data on their activities (Helmond, 2015). Initially concentrated in the domain of social media, the word platformization has come to refer in the literature to the spread of the platform model ‘into the rest of the web’ (Helmond, 2015: p. 1), with increasing influence in economic, cultural and governmental fields (Van Dijck et al., 2018).
One of these spheres is employment, with scholars employing the term platform work (also known as gig work) to describe diverse forms of employment ‘mediated through platforms’ (Graham and Woodcock, 2019: p. 3). Some of this work has clarified that platforms are not simply changing the way people look for work but working to shift the very nature of work – so that it becomes platformized. For instance, Gandini (2019) argues that through intermediating a tripartite relation (worker-consumer-platform operator, see Ződi and Török, 2021) the platform becomes the place of employment, containing all the social relations inherent to the production process. In this way, platformized work un-fixes employment relations and fences off production relations ‘in a newly delimited, unique environment’ (Gandini, 2019: p. 1040). Hence, gig work platforms create new walled gardens which restrict and control user interactions, whilst harvesting valuable data from both consumers and workers (Chen and Sun, 2020; van Doorn and Chen, 2021). This allows them to produce ‘network effects’, whereby ‘the more numerous the users who interact on a platform, the more valuable the entire platform becomes’ (Srnicek, 2017: p. 43), because greater numbers of users mean more data and more accurate predictive algorithms, which makes it more attractive to users in turn. In summary, the larger and more established platforms become, the more powerful they become and the more agency they can exert.
This brings us to another key effect of the platformization of work and a key point of agency for the businesses that operate platforms: algorithmic management. Data extracted from every interaction on the platform is fed into artificially intelligent algorithms, which then attempt to manipulate the worker using gamified incentives, into making themselves available for work at the places and times where they are needed most, whilst increasing their productivity and, thus, profitability for the platform companies (Ravenelle, 2019; Stark and Pais, 2021; van Doorn and Chen, 2021). At the same time, algorithmic management has (until recently) largely escaped recognition as management by lawmakers, which has allowed platform operators to continue their business model based on them being classed as technological mediators who merely connect to independent contractor/partners, rather than employees, the latter of which have much greater risks and liabilities (Gandini, 2019; Graham, 2020; Koonse et al., 2021).
However, scholars have criticized the portrayal of algorithmic management and the platformization of work more generally as producing even effects, no matter the context, as being overly linear and deterministic (Bissell, 2023; Gregory and Maldonado, 2020; Moisander et al., 2018). Indeed, Bissell (2023: p. 61) urges the need to avoid totalizing claims around the transformative nature of platforms, and the need to pay closer attention to ‘gaps, aporias and absences’, as well as the complexity and unpredictability of local contexts.
In line with this focus on the unpredictability of effects, Niels van Doorn and colleagues at the University of Amsterdam have argued for studies on the platformization of work that pay closer attention to variability – how things ‘often do not go as planned’ (Van Doorn et al., 2021: p. 715). ‘AEP’ (Van Doorn and Bosma, 2021) is a granular, case-by-case focus on the ways that platforms seek to embed themselves, creating and exploiting partnerships in different settings as a way to ‘expand their multisided markets and gain sociopolitical influence’ (van Doorn et al., 2021: p. 726). This reliance on partnerships draws connections to Srnicek's (2017) earlier emphasis on platforms as intermediaries – with their agency primarily deriving from their ability to connect different agents (other companies, governments, workers, consumers, etc.) through the affordances of the app and the data generated through user interactions. The model is reliant on owning minimal fixed physical assets – thereby remaining disembedded from physical locations and the associated risks and liabilities (Graham, 2020). However, it cannot escape the physical world entirely, creating ‘material, situated dependencies’ (Van Doorn et al., 2021: p. 715) that place constraints and limitations on the agency of platform operators, leaving the process of platformization more open to the agency of others (human and non-human).
Geographers of neoliberalization (Brenner and Theodore, 2002; Peck, 2010) have long called attention to the disjunct between the seemingly neat ideology of free markets and the ‘messy hybrids’ (Peck, 2010: p. 24) and compromises that result in particular national and regional contexts. Indeed, van Doorn et al. (2021) drew on Brenner and Theodore's (2002) call to study ‘actually existing neoliberalism’. Like neoliberalization, this article argues that platformization should be understood as ‘produced within national, regional, and local contexts defined by the legacies of inherited institutional frameworks, policy regimes, regulatory practices, and political struggles’ (Brenner and Theodore, 2002: p. 349), which can create ‘backflows, and undercurrents’ (Peck, 2010: p. iii). It is workers’ political struggles and the resulting agency that we want to concentrate on next, as we argue that this aspect was underdeveloped in van Doorn et al.'s account.
Worker agency and platformization
Scholars have noted that platform work ecosystems can be coercive and isolating environments (Koonse et al., 2021; Rosenblat and Stark, 2016), antithetical to the formation of collective identities and agency-creating collective action (Anwar and Graham, 2020; Salter and Dutta, 2022). However, they can also open up spaces for resistance and therefore the ability to exert agency in the shaping of platformization. Subsequently, scholarship has begun to recognize and theorize the emergence of collective action such as strikes and protests (Cant, 2019; Cant and Woodcock, 2021; Howson et al., 2020; Woodcock, 2021), as well as novel communicative tactics such as WhatsApp groups and other forms of social media organizing (Cant, 2019; Grohmann et al., 2023).
However, much of the scholarship on platform worker agency has thus far been in the male-dominated areas of rideshare and delivery (reflecting the majority of scholarship on platform work). With some notable recent exceptions (Altenried and Niebler, 2024; see Khan et al., 2024; Ticona and Tsapatsaris, 2023), there has been a dearth of work which focuses on the kinds of agency that accumulates in women-dominated platform-work environments such as care work, where informal networks of solidarity and support have been long cultivated (Bansal and Arora, 2023). Resistance and agency in such contexts can employ the more full-frontal tactics of strikes and protests (Anwar and Graham, 2020; Salter and Dutta, 2023) but also encompass more subtle and informal articulations of a feminist ethics of care (Khan et al., 2024; Mountz et al., 2015; Tronto, 1998).
Care work is inherently relational and empathetic, meaning it is difficult for care workers to prioritize their material interests, creating a conflict with the preferred competitive, entrepreneurial identity for platform workers (Khan et al., 2024). As argued by Khan et al. (2024), an ethics of care can not only create bonds between carers and the cared for but also between workers, who form solidarity ties around a shared focus on prioritizing care for patients over their interests or those of the company. As Moore and Hayes (2018) argue, the complex relational aspects of care work inherently resist being broken down into piecemeal, disaggregated and standardized task lists to be ticked off on an app. Research in other care environments highlights that for care workers, resistance can include a resistance to hierarchy, where care workers query the capacities of the senior staff (Lee-Treweek, 1997), and resistance to the standardization of emotional labour (Hochschild, 2012), by engaging in emotional work outside of the ‘professionalized’ care relationship (including treating care recipients as family rather than as clients).
The kinds of feminized, micro, everyday resistances through discursive and physical care practices may not at first appear to challenge the business model of platform capitalism. They can be trivialized through misogynistic terms such as ‘gossip’ (see Ticona and Tsapatsaris, 2023). However, when collectivized through informal networks, they can exert agency through challenging and pushing back against dominant narratives and ideologies (Chen and Phan, 2023; Grohmann et al., 2023; Ticona and Tsapatsaris, 2023). This agency can often manifest itself through simply using the platforms in ways that are unintended by the designers, through which they work to ‘configure[e] the meaning and uses of the technologies’ (Lupton, 2017: p. 12). In Peck's (2010) terms, workers can create ‘backflows, and undercurrents’ (p. iii) which can work to shift platformization from a neat, linear path.
As we explore in the analysis and conclusion sections, we therefore argue for a scholarship to place close attention to two levels of worker agency in shaping the path of platformization in any given context: the formal level of collective union representation as well as the informal level of everyday uses/misuses of technology and unpremeditated networks of solidarity. However, firstly we outline how spatial factors can exert agency and shape the path of platformization.
Spatiality as an agent in the rural context
The AEP perspective (see Shelton and Lodato, 2019; Söderström and Mermet, 2020; Van Doorn et al., 2021; van Doorn and Bosma, 2021), draws on a geographical tradition that focuses attention towards ‘space and place’. In Aotearoa, through the Indigenous Māori culture, there is a long history of considering space and place as agentic, and Kaupapa Māori research has begun to trace the similarities between new materialist, socio-materialist and Indigenous approaches, which challenge Western assumptions of human/object and mind/world dualisms (Moewaka Barnes and McCreanor, 2019; Rua et al., 2023). For the precolonization Indigenous Māori of Aotearoa, the land (whenua) could not be separated from the people (as well as their ancestors) who collectively occupy and use it (hence Māori are often collectively described as mana whenua – people of the land). Western ideas of individual ownership of land and its commodification to extract maximum short-term profit were alien to Māori precolonization, as well as the separation of the health of the land from the health and well-being of the people (Moewaka Barnes and McCreanor, 2019).
As well as being inseparable from the human, the land (whenua) and other spatial phenomena (including mountains (maunga), rivers (awa) and the sea (moana)) ‘possess their own wairua (spirit, sense or soul) and agency’ (Rua et al., 2023: p. 41, emphasis added), rather than simply being inanimate objects to be acted upon by humans. Rua et al. (2023: p. 44) emphasize how this Indigenous stress on the non-human agency of spatial factors aligns well with recent shifts in Western social theory, which is beginning to acknowledge the human subject as ‘a nexus of interacting assemblages that are reproduced through ongoing social relations, human and non-human interactions and institutional practices’.
New materialist and socio-materialist theory has been recently utilized in studies of platformization, focusing on the material agency of technology and data (see, e.g., Aasback, 2022 on social work; Westermeier, 2020 on financial transactions). A significant body of further work has explored the environmental impacts of technology production and data infrastructures (see, e.g., Parks and Starosielski, 2015). However platformization scholarship's thus far predominant focus has been on the urban, over the rural, emblemized in the term ‘platform urbanism’ (Barns, 2020; Graham, 2020), which in recent years has become its own sub-discipline (Bissell, 2023). Indeed, Van Doorn et al. (2021), in their conceptualization of AEP, understand it as ‘a predominantly urban phenomenon’ (p. 728), a focus which neglects rural and semi-urban contexts.
Pushing back against this trend, scholarship in the Chinese context has called for attention to be placed towards ‘platform ruralism’, through an increased understanding of ‘the unfolding of digital platforms in rural areas’ (Wang et al., 2022: p. 12). Similarly, Zhang and Webster (2024) ‘argue for greater attention to the role of rural space in platform geographies’ (p. 131), and the ways that rural contexts do not always mirror the urban, with local (often more traditional) norms and structures shaping how the platforms become ‘user-fied’ (the negotiation of meanings around the technology and its subsequent uses).
However, such studies of rural contexts have not so far specifically considered the agency of spatiality; specifically, the ways that mountainous terrain can shape how connectivity unfolds. Limited cellphone and internet network coverage can create interruptions and disconnections which influence the trajectory of platformization. We will explore how this has unfolded in the Aotearoa context in the below analysis. However, firstly, we outline our method and methodology.
Method and methodology
Two methods were employed in this study: interviews and document analysis. 16 interviews and one focus group were conducted with Aotearoa-based care workers between February and November 2022. Participants were asked to complete a demographics survey before the interview, the results of which are displayed in Table 1 below (note – not all of them did this, which is why Nawal's age is missing).
Interview participant demographics.
Whilst the sample size is too small to claim any statistical inference, as shown in Table 1 above, 14 of the 16 participants identified as female (87.5%) and 2 male (12.5%). In terms of ethnicity, 14 identified as Pākehā (the Māori word for white European setter) and 2 Asian (both had migrated to Aotearoa). We did attempt to recruit more migrant workers through placing an advertisement in Migrant News, but it failed as a method for reaching this group.
In terms of age, two were aged 24–34 (13.33%), two aged 35–44 (13.33%), four aged 44–54 (26.66%), two aged 55–64 (13.33%) and five aged 65–74 (33.33%). Whilst over 30% of participants being over 65 was an outlier in comparison to the wider population of care workers (see Ravenswood et al., 2021), the age distribution is generally reflective of the Aotearoa care workforce – particularly the low percentage of workers under 44.
Geographical distribution is also worth noting, due to its relevance to our argument on the importance of spatial factors and paying attention to the rural context. Just 7 of the 16 participants (43.75%) were located in one of Aotearoa's six major urban centres, which are defined as having populations over 100,000. The other nine were either located in smaller regional urban centres (which have strong links to their rural surroundings) or small semi-rural towns, where many of their clients would be located rurally. Four participants were located in the South Island (which is larger in geographic size but smaller in population than the North Island and much more mountainous).
Following ethical approval, recruitment proceeded through the sharing of a poster on a closed union Facebook Group, facilitated by contacts. Due to the pandemic context and the geographic dispersal of participants (located both in major cities and rural locations across both of Aotearoa's major islands), the interviews and focus group were held on and auto-transcribed using Zoom. Transcripts were then checked for accuracy by a Research Assistant. Interviews were in-depth and semi-structured, lasting between one and two hours. Whilst a schedule of themes was followed, this remained flexible enough to allow for the articulation of anecdotes, which were often not only highly insightful but also aided the building of rapport, which is important for ensuring participants are comfortable with probing and politically loaded questions (Johnson, 2002; Tracy, 2019).
Further, a document analysis (Bowen, 2009) was conducted of a corpus (n = 67) of policy documents, web pages and media articles related to two care platforms which we argue are representative of those that have entered the Aotearoa market: Mycare and My Homecare. Document analysis was selected as it is ‘particularly applicable to qualitative case studies’ (Bowen, 2009: p. 29; see also Howarth, 2005), enabling triangulation with and comparison to interview data.
Once both sets of data had been collected, thematic analysis (Braun & Clarke, 2006) was employed, where initial descriptive codes were generated before merging these into broader, theoretically informed themes (see also Tracy, 2019). Both authors initially coded each transcript individually, keeping separate codebooks, which were later discussed before agreeing on the most prominent themes. This was facilitated through a shared folder in the Dropbox platform. Quotations from the interview transcripts are reproduced below, attributed to pseudonyms. Employers and work locations are also withheld for confidentiality.
The following two results sections outline the four key themes (the four factors described earlier), which we have theorized as agencies that influence the platformization of the home care sector in Aotearoa: policy, platform, workers and spatiality.
Curtailed agency: a comparison of two Aotearoa care platforms
Mycare
Mycare's path of platformization has been significantly shaped by Aotearoa's home care policy settings. As detailed, care work platforms which hire workers as independent contractors, which have become highly prevalent in many parts of the world, are not yet widespread in Aotearoa, because of a particular and complex funding regime, itself a happy accident of four decades of neoliberalization. One exception is Mycare, a platform that connects independent contractor care workers to elderly clients who can afford to go private, as well as those with disabilities from the public sector. Mycare matches care workers (who curate profiles) to clients and provides the latter with administrative services, such as criminal record checking and the negotiation of pay rates and contracts. For workers, the platform also provides time scheduling and tax support. As with care work platforms found overseas, workers on Mycare are rated by the client with a 1–5 star rating (as well as having the ability to leave written reviews), and a low rating influences workers’ ability to find sufficient, well-paid work in the future. Further, workers are algorithmically matched to clients according to details that they enter on their profiles, such as location and experience.
The platform launched in 2013 as a non-profit, describing itself as a ‘social impact platform company’, with a number of Aotearoa's most influential business people coming quickly on board as investors and/or directors (My Homecare, 2018). Mycare's website (2018) explicitly states that they are ‘not a provider of home care or support services’, thus allowing them to sidestep the regulations which come with directly employing care workers. In line with the ‘social impact platform company’ brand, and to sidestep a funding regime which prevents them from accessing public money for elderly care, the platform has diversified by investing in Te Heke Mai; an employment readiness and coaching service delivered via an app in partnership with the Ministry for Social Development (Te Heke Mai, 2019).
Thus, through their institutional connections and partnerships, Mycare is embedding itself in both the health and social sectors of Aotearoa, enacting platform agency in both spheres. However, through positioning themselves as a mediator rather than a provider, they are also sidestepping regulations that govern how traditional care providers operate, such as a legal requirement to directly employ workers, shielding themselves from some of the effects of both policy and spatiality. Because the care workers who use the Mycare platform are contractors, rather than workers, they lack the ability to join unions and formally organize or bargain, placing certain constraints on formal worker agency.
Conversely, however, Mycare's small size, due to their lack of access to public money for elderly home care (which makes up the majority of the home care funding in Aotearoa), limits their ability to collect enough data to build network effects through the increasingly accurate algorithmic management of workers. At the same time, the two unions that represent care workers in Aotearoa, E tū and the Public Service Association, are aware of the threat that Mycare's model poses and have had several years to campaign, organize and educate against it (see McCully, 2021).
Nonetheless, Mycare is positioning itself to benefit from a possible future funding regime change (where they gain access to public money for elderly home care) by linking their technology with the agency of those who need care, through the discourse of choice, empowerment and autonomy. An article on their blog titled ‘Empowerment and autonomy’ (Brunzel, 2023b), links the platform with challenging ‘able-bodied people's perspectives’ on disability through giving voice to the disabled'. Whilst another, from a carer's perspective (Brunzel, 2023a), draws links between the platform and the return to community-based care, where otherwise socially isolated elderly people in rural locations can connect with eager carers in their local community. Hence, the technology is framed as allowing the sidestepping of traditional, bureaucratic large provider-based care (and implicitly the unions associated with it), which would happily see the ‘health dollar gobbled up with travel expenses', by assigning ‘an agency carer [to] drive for over an hour away'. At the same time, Mycare's website provides the means for care clients to enact agency through writing to their local DHBs (who Mycare is also ‘in conversation with’) to request control over the way their public funding allocation is spent, making a future change to the funding regime more likely.
My Homecare
More recently, one of the largest care providers in Aotearoa, Geneva, has developed My Homecare (2018). This platform incorporated some of the technological architecture of care work platforms like Mycare, including the ability for ‘clients to pick and choose carers based on personal profiles’ (Slade, 2018). Accordingly, the website echoes Mycare's discourse of personal choice, empowerment and autonomy – claiming to offer clients the ‘real choice’ over which staff work with them and when (Geneva Healthcare, 2018). Hence, the development of this platform in 2018 represented Geneva both looking to compete with Mycare for the private care market and looking forward to an anticipated policy move towards consumer-led, individualized public funding (Slade, 2018). In the meantime, Geneva is creating dependencies on their platform infrastructure for public funding bodies that privilege efficiency and are easily seduced by seemingly easy technological fixes, whilst being long dependent on the infrastructure of private providers.
However, for the time being, their main business is the provision of elderly care via public bodies that allocate their funding based on a quota, rather than individual clients having full control over the distribution of their care funding (as in other countries such as Australia). As described, this means that Geneva must employ care workers directly, with the majority being on permanent contracts with annual leave, sick pay and other employment rights. This means they have been unable to disembed themselves from the local policy settings, worker agency and spatiality to the same extent as Mycare.
We argue that these forms of agency have impacted both the My Homecare platform, which is the portal for clients to liaise with workers and Geneva, as well as the Geneva LIVE Mobile App. This app mediates between workers and Geneva, with the former required to download it onto their personal phones. As part of Geneva's publicity, it purports to offer accurate and timely pay, ‘instant access to your past, present and future schedule’, ‘the best travel routes with Google and Apple Map’, ‘easy communication with Geneva’, the ability to apply for leave and ‘a comprehensive overview’ of client details (Geneva Healthcare, 2023). However, as we outline in the below section, according to the workers who used them, many of these functions did not work, adding additional unpaid labour and digital frustration, rather than empowerment.
From the outset, the agentive potential of the Geneva LIVE app to platformize care work in Aotearoa was recognized by the unions (with Geneva being Aotearoa's largest home care provider), who thereby ensured its rollout became a site of fierce contest. Unlike Mycare, the first pool of care workers to use it was not individual contractors but instead part of a highly organized and unionized workforce, which demonstrated formal worker agency through articulating a critical collective voice, including writing an open letter to the Privacy Commissioner to express concern about the types of digital information that might be collected about them, as well as articulating concerns around declining working conditions such as the erosion of guaranteed hours (E tū, n.d.). The two unions were also strategic in voicing their criticisms through the media, with E tū attaching the catchy label the ‘Tinder of homecare’ to the platform (Slade, 2018), with the metaphor doing important discursive work in connoting broader concerns, held by both care workers and the public, around dehumanization and the erosion of the relational aspects of care work.
Geneva strongly and publicly rejected these claims, reframing the discourse around the app in terms of client rights, who themselves were represented as lacking a voice, as well as ‘vulnerable and in need of support’ (Geneva Healthcare, 2018), thereby denied the agency to have control over their care. In a similar way to Mycare, this rhetoric, which was carried into the news media, attempted to reframe the concerns of care workers into a focus on client autonomy enabled by a ‘technology solution’. Despite the attempt to shift the conversation, the success of the app ultimately relied on access to qualified, experienced workers, and Geneva was forced to acknowledge their concerns more directly. Late in 2018, the two unions, along with Geneva, released a set of guidelines related to My Homecare (E tū, 2018), designed to ensure privacy (workers were entitled to access their data) and working conditions (the legal requirement for care workers to be paid a number of guaranteed hours each week was to remain protected). Since 2018, seeing the ‘efficiencies’ Geneva has been able to introduce (e.g. firing management and administrative staff in local offices), most of the other providers in Aotearoa have also introduced similar care apps, with varied degrees of complexity, functionality and efficiency. The effects of this shift on the working conditions of the care workers we spoke to will be discussed below.
Working with apps: contradictions, undercurrents and backflows
The path of platformization and worker experiences have been significantly shaped by the agencies of the non-human: notably the material functioning of home care apps and Aotearoa's specific geography. What became clear during the interviews with care workers was that whilst the introduction of the care apps did enact agency in significantly impacting working conditions, these effects were variegated and marked by contradictions, undercurrents and backflows. Spatiality emerged as a significant factor. Further, the use of the apps was marked by technological and communicative breakdowns, creating significant inefficiencies with resultant frustrations at the (mostly unpaid) additional labour requirements and risks.
The geographic environment of Aotearoa coupled with policy and commercial decisions related to communications infrastructure shaped how the apps worked in practice. Mountainous terrain often makes rural locations quite remote and difficult to access, with limited cell and internet network coverage. This meant that functionality central to the imaginary of platform care (such as tracking workers using Global Positioning System or using an automated time sheets through the app) simply did not work. For some participants who worked in rural or semi-rural areas, the apps would often simply not function at all. Workers in these regions were highly autonomous insofar as they had little communication with management. In Aotearoa, some home support providers offer employees cell phones (other providers, notably Geneva, position a self-provided smartphone as a condition for work). Poor cell phone coverage meant that some carers could not use the smartphones provided by their employers. Further, because pay slips, time sheets and other administrative matters were mediated through the app, when the app ceased to work because of spatial factors, the effect was akin to ‘rolled back’ employment protections.
Because of the unreliability of the use of apps in the Aotearoa context, platformization was partial, marked by contradictions and heterogeneous systems. Contact between workers, clients and providers occurred in highly variable technological configurations, from the relatively sophisticated apps to the surprisingly basic. Scanned printed documents sent via email, digital photos, text messaging and phone calls remain some of the ways run sheets are delivered, time sheets logged, client information provided, and notes taken. Whilst there were spaces for workers to exert agency, this piecemeal, heterogeneous system overwhelmingly functioned to benefit the provider-employers.
For the workers we spoke to, the apps were perceived as designed to maximize economic returns at the expense of working conditions. Because of the progressive degradation of their workplace conditions and entitlements over the years, the home support workers we spoke to were highly sceptical about the promises of care apps, contesting Geneva's discourse of digital convenience and empowerment. By contrast, participants using the apps described their impact as marked by inefficiency, precaritization and dehumanization. As Sarah describes below, replacing local managers with a call centre meant both the loss of contextual knowledge about the clients and the loss of clinical expertize: the restructuring, it's had a big impact on how we work. We had a manager in our area in the [local] hospital that we could go and see, she knew all our clients. […] they changed that and made it a call centre. While I really respect the people in the call centre, they have no idea […] They have no background or training like we have…
The above quote demonstrates how the imposition of a call-centre structure, as part of the process of platformization, does enact agency in placing certain constraints on communication with management, which becomes largely mediated through the smartphone, with face-to-face contact becoming increasingly rare. Using apps and call centres to mediate between workers and employers implicitly changes the nature of employment relationships in a way that makes realizing full workplace entitlements harder, becoming more akin to gig workers/contractors on an informal level, whilst still retaining the formal legal status of employees.
Many of our interview participants spoke of the difficulty of being paid correctly, which added to already precarious working conditions. As mentioned, this was compounded by spatial factors – logging in and out was not possible for some in rural settings, meaning confirmations of their visits to clients were often inaccurate. Further, the functionality of the care apps made managing workplace entitlements harder. Contrary to Geneva's official communications, according to the workers their platform did not store long records of past rosters and jobs worked, so actually checking pay was hard. Whilst this is partially explained by the fact that adding in extra clients at the last minute is a frequent practice, in theory, the app should make this easy to administer. Whilst the introduction of the app meant that the process of applying for leave was streamlined, many participants related how no automated record of applying for leave was generated, meaning requests submitted were frequently ignored. Workers would then have to spend significant time and energy following up the request. Here, spatial, technical and commercial actors interact to shape platformization in a way that rolls back workplace conditions.
AEP is shaped by these actors in ways that impact the ‘efficiency’ of care, notably in ways that impact workers and clients. One effect of app dysfunction is that since logging in to electronically record jobs is not possible, it becomes even more difficult to claim overtime or for workers to check they were paid accurately. Due to spatial factors, app-based directions and Google Maps do not always provide the fastest, most efficient route between clients. The rosters are created by office staff located in urban centres with little to no knowledge of these rural areas. Hence, the system of having rosters created centrally and delivered electronically creates efficiencies in saving the company money, whilst encoding inefficiencies for workers into the app.
The combined effect of the care app interface and the rationalization of the office staff is one of inequities, contradictions and failures in information flow that characterizes the platformization of care in this context. The workers we spoke to lacked access to all the information they needed to do their jobs safely, whilst at the same time generating large amounts of data through their use of the apps which can be utilized by providers for profit. Participants regularly reported that the information they were given about clients was frequently out of date or erroneous, meaning they were unable to exert agency over the types of work they accepted.
Functionality that could be beneficial for care was often not part of the platformization process. Workers reported having significant difficulty communicating serious problems with clients to management, adding to a perception that the motivation for app-based technologies with tracking functionality appears to be more about finding rationales for not paying workers for time segments. Compounding this, many workers described a frustrating situation where observations about clients’ deteriorating health, circumstances or home environment were routinely ignored.
Such incidences are profoundly disempowering and prevent care workers from being able to offer quality care. To cope with such situations, care workers were reliant on informal networks, sharing information about clients amongst themselves and creating spaces for debriefing after traumatic incidents. In this quote, an interviewee describes how she is part of a local support network of care workers that used Facebook Messenger to keep in touch, following the closure of the local offices and the move to the call-centre model: …we just say stuff, I mean we have a messenger line between us […] that was definitely one for all and all for one […]. It's the Service Center […] the centralization program […] and the hell bent on having everything run from […] a central point and with coordinators who really have no idea about what goes on…
The interviewee is describing how the support network fostered an empowering sense of collectivity – emblemized by ‘one for all and all for one’, which pushed back against the otherwise isolating and disempowering effects of platformization. Workers can therefore use informal communication channels to exert agency in shaping platformization. This is particularly well highlighted in this next quote which details how they prioritize an ethics of care over reliably logging into the app at the correct times, disrupting reliable data collection: Quite frankly I just don’t care about the app, if I forget to log in, I don’t care. I log in when I feel like it. I’ll log in half way home. To me, it's about the client, it's not about my phone being more important than the cares.
Informal networks became a particularly important resource during the COVID pandemic, which created significant problems for care providers with ‘lean’ structures that, compounded by government failures to procure supplies, struggled to distribute personal protective equipment. This interview participant recounted, at this time when the health risks associated with information breakdowns and inefficiencies were accentuated, how a local support network compensated: there was about six of us and we all swapped stuff that we needed […] high need clients so I needed PPE gear. So you know, I scored a box of gloves from one support worker and [laughter].
Whilst another described how they monitored the union support worker Facebook group to answer questions from other workers, in their time, filling in information gaps left by the provider: I do go on the home support worker's group. I like to see what's going on at other companies and what's going on with girls and maybe try and answer any questions that they've got where I can. If it's [employer] related then I can answer questions for them, lead them in the right direction.
In these examples, informal support networks are a way for care workers to deal with their negative working conditions and find ways to enact agency in retaining some sense of control. However, beyond simply providing an outlet for specific practical support, informal networks were a site which foregrounds recognizing, responding to and materially addressing the complex needs of clients and other workers (Tronto, 1998). Platformization minimizes the relational aspects of the job, embedding granular, processual meanings into support work. However, through informal support networks, holistic information about clients, their life worlds and their needs is shared between workers, enabling them to identify needs in others and themselves and respond in material (sharing equipment) and immaterial (sharing knowledge and experiences) ways. Structural pressures to become monadic, self-reliant entrepreneurs were resisted by most participants through maintaining and strengthening informal support networks. The resultant micro-resistive discursive practices shape how platformization is realized on the ground.
An ethic of care that prioritizes collective well-being, sustaining both workers and clients (Khan et al., 2024; Mountz et al., 2015; Tronto, 1998), over the pressures of techno-capitalist efficiencies, articulates an alternate identity for ‘care worker’ that feeds into larger, union networks. Here, collective action at the informal and formal levels has a structural impact in shaping care policy and workplace conditions. In this regard, workers exercise significant agency over the structural path of platformization.
Conclusion
This article, through an analysis of the home care sector in Aotearoa, has outlined four key interlinked sites of agency that have shaped the path of platformization: policy settings, platform operators, workers and spatiality. All of these are complex sites of interactions between a host of human and non-human actors. Whilst this typology has relevance to the platform and platformization studies more broadly, our key focus here has been building on and developing van Doorn et al.'s (2021) conception of ‘AEP’, arguing for more scholarly attention towards worker agency (both the formal and informal levels) and spatial factors (which is particularly important in rural contexts). We hope that this typology, in urging scholarly attention to broad sources of agency, some of which are less visible than others, will be useful for future research in the platformization of work.
This article thereby argues that there is both the need to chart ‘the institutional boundary work achieved through partnership building’ (Van Doorn et al., 2021: p. 727) by platform operators, and how the often-chaotic consequences of these moves are interpreted, coped with and resisted by workers. Whilst provider-led platformization did enact significant agency in implicitly shifting the nature of employment relations through mediating worker communication with management, the workers also generated significant agency in pushing back against this, both at the formal level of unions and through informal networks, which also allowed them to cope with the disruption and decline in working conditions through a shared articulation of a feminist ethics of care.
Combined with worker agency, Aotearoa's unique policy settings have also produced unexpected contradictions and backflows, blocking the linear flow of platformization, so far limiting the ability of Mycare to gain a substantial foothold and meaning Geneva LIVE could not implement all of the functions of a care work app for the majority of its users. The agency they have managed to assert has been via partnerships with powerful state and private actors, meaning they are playing a longer agentive game; putting pressure on and creating the discursive conditions for future funding regime change, whilst in the meantime ingraining dependencies in a system reliant on private provision.
Aotearoa's unique spatiality, in conjunction with policy settings, also shaped the path of technological development in unexpected ways. Mountainous terrain and the uneven spread of populations often limited cell phone coverage, curtailing the effectiveness of certain app functions, such as maps and automated location tracking of care workers. Whilst this could curtail the agency of workers in presenting frustrating barriers and even severe health risks, it also presented more opportunities for the shared articulation of resistive care ethics between workers.
We acknowledge that this study has limitations in illuminating the full complexities of the platformization of home-based care at a global level. As we have outlined, Aotearoa is a unique context, in terms of its small population, rugged terrain and unique policy settings, the findings from which may well not be applicable in other settings. However, this uniqueness is also its strength, representing a call for future research to ensure that all complexities are accounted for in a given context.
Another distinctive feature is the strong presence of Indigenous Māori culture, which sits side by side with the dominant postcolonial European settler culture. Hence, one limitation of this study is that we were unable to represent this biculturalism more fully, aside from a mention in the literature review. The study drew on the perspectives of mostly Pākehā interview participants and authors. Future research in the context of Aotearoa could draw upon Kaupapa Māori theories (see Hikuroa, 2018; Mahuika, 2008; Pihama et al., 2015) to explore ‘actually existing’ research from an Indigenous perspective.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the Ministry of Business, Innovation and Employment (grant number MAU2102).