Abstract
The proliferation of digital labour platforms ranges from ride-hailing and delivery services to care work. Eldercare platforms emerged as a prominent focus of critique regarding platformization, commodification, and the nature of care itself. However, empirical studies thus far have largely focused on domestic work and childcare, leaving a significant gap in the examination of eldercare platforms. This paper addresses this underexplored area by presenting empirical findings derived from a mixed-methods analysis involving 2144 eldercare workers’ profiles and 14 in-depth interviews. Through cluster analysis, the research identifies the distinct groups that comprise the workforce on on-demand eldercare platforms. In doing so, the study challenges prevailing narratives around platform labour, emphasizing that they are not universally applicable. The findings reveal that platforms tailored for cleaning, domestic care and eldercare each engage uniquely composed workforces, highlighting the significant stratification within care work and its implications for platformization. Ultimately, the research presents three key insights: first, that care platforms function as niche markets; second, that workers in these on-demand platforms possess substantial practical experience and qualifications in caregiving; and third, that, contrary to trends in other sectors of platformized care, migrant workers are not disproportionately represented on eldercare platforms in Germany.
Keywords
Introduction
Labour scholars and journalists alike have expressed major unease about labour platforms penetrating the field of caretaking and care labour. While some acknowledge the opportunities this may provide for the elderly and care provision, such as being enabled to stay at home (Wörle et al., 2023), most scholars emphasize the fact that labour platforms foster precarization of workers, especially the most vulnerable female and migrant care workers (e.g. Pulignano et al., 2023). After a period of empirical underrepresentation of care platforms in comparison to other types of digital labour platforms, studies on platformized forms of care have surged. Despite this recent development, the field of eldercare and especially the perspective of eldercare workers using platforms to find clients remains largely unexplored (Orth and Baum, 2024).
This is particularly surprising for three reasons: First, care crisis arguments tend to stress the risks platforms could mean particularly in eldercare work (Dowling, 2021, 2022: 6; Strüver, 2021). Second, the field of long-term care (LTC) poses the most prevalent care gap within Western welfare states (Dowling, 2022; Trojansky, 2020). Third, the policy question of how to support the elderly at home is especially salient within the European Union (EU), the region that is the empirical focus of this paper.
Moreover, LTC is a field where digital and smart technologies dominate the search for solutions in handling the void of care and workers (Hoff and Pottharst, 2023). Platforms for eldercare have thus become a symbol for discussing the negative outcomes of digitalizing welfare states. Given the importance of this field for academic knowledge, policy and practice, it is highly relevant to carry out empirical work to complement and expand on the many conceptional studies on care platforms, commodification and crisis of care. This paper therefore provides an empirical case study from within the EU, concretely Germany, into workers accounts on eldercare self-employed and platform work, centring on workers who use an on-demand platform and workers who deliberately choose self-employment 1 as routes into the care vocation.
Scholars have repeatedly shown that eldercare and care work platforms function in a specific ecosystem of national and supranational care regimes and platforms can be best understood as part of this ecosystem. Taking the industry specifics into account means considering not only the specifics of the tasks and of home-based services. Rather, it is also crucial to consider that care workers are scarce, demand for daily support is extremely high (Hoff and Pottharst, 2023), and that marketized cash-for-care funds are available but thus far underutilized (Scheerbaum et al., 2024).
For this reason, it is important to empirically research those who offer work on platforms and are self-employed in the field of eldercare. To situate the empirical material, the paper analyses the role of the welfare state environment and the question of stratification of care workers within the care ecosystem of different and new intermediates from more traditional agencies to platforms and facilitators for daily support. Then, the paper uses original data from a 5-year investigation into care platforms in Germany. The paper draws on a mixed-method design with two types of data: The first is a set of worker profile data, gathered through a web crawl. It contains the information of 2144 individual care workers registered with a German care platform. To corroborate these data, the analysis also draws on 14 narrative interviews with care workers. By doing so, this article not only centres on an on-demand platform and care work. It also hones in on the general question of self-employment in care, when, how and for whom self-employment (and platforms) have become a valid option and why that is. This is crucial as previous findings show that eldercare workers do have multiple options of intermediation (Rodríguez-Modroño, 2023, 2024) and might also gain autonomy and agency through self-employment in a care ecosystem where marketization seems omnipresent (Baum and Kufner, 2021; Muldoon et al., 2024).
The analysis in this paper begins with a literature review on care platforms, including the context of marketization of care in the EU, and introduces the context of the German welfare state to situate the empirical material. It then offers a brief description of the research design and methods applied. Following from this, I present the empirical findings, which are structured along five dimensions: (1) general demographics of care workers on the platform, including age, regional origin, potential migration status; (2) qualification and occupational groups, experience and prevalence of types of workers in each cluster formed; (3) pay rates and availability; (4) offered tasks; and (5) motivations and orientations. In the discussion section, the findings are contextualized with prior findings and the limitations will be spelt out. The concluding section includes the three main findings from the empirical study.
In sum, the findings corroborate previous research from Spain (Rodríguez-Modroño, 2024) and Australia (Macdonald, 2021; McDonald et al., 2021, 2023), in that eldercare is reorganized in the advent of digitalization and care crisis. However, it also shows continuance, such as underpaid care workers taking on additional jobs in home care. Moreover, findings confirm mentioned studies with regard to major stratification among care workers and that the welfare state plays a crucial role for platforms and for workers in that it shapes the ecosystem of care on a national and supranational – in this case EU – level.
Literature Review
This literature review focuses on platform studies and provides various arguments for how domestic and care platforms are specific. It offers an overview of the features of care platforms thus far carved out by researchers. In the first part, it shows how care platforms are commonly defined and set apart from other forms of platform labour. In addition, it presents findings from specific sub-fields of care, namely childcare, domestic workers and cleaners. The following parts of this section focuses on findings from studies on eldercare. As this review will show, there are few insights thus far in the field of care platform studies that build on the experience of eldercare workers. However, recently, authors have argued that the welfare state ecosystem is of great importance and that within eldercare there seems to be a specific stratification among different agency models.
Platform Typologies and Defining Features
Scholars have repeatedly shown that labour platforms form an organizational model that may restructure work processes with negative outcomes for the workers. In short, according to Gregory (2022), ‘numerous studies now show, this type of work is bad for workers across several markers from financial to psychological’ (p. 925). Commonly, platform labour is divided into on-site and online work and marketplace and on-demand labour platforms, as this is associated with different labour processes and implications for workers, such as pay, employment/contract matters, autonomy and control (De Groen et al., 2018). A gender divide in platform studies, however, means that findings are often based on male-dominated occupations (Grau-Sarabia and Fuster-Morell, 2021). Nevertheless, a vast range of work tasks can be found in the literature on platform labour (Schor et al., 2020). These can be paid and unpaid and also reproductive, such as content moderation or ghost work (Jarrett, 2016). Studies have scrutinized three major strands within platform labour research: (1) misclassification and specific exploitation of workers, (2) questions of access for low-skill and migrant workers and (3) the enhancement of precarity (Cornet et al., 2022; Schor et al., 2020). These common understandings of the problems within the research field have, however, been frequently complicated by scholars focusing on care (Ticona and Mateescu, 2018).
These critiques centre in the fact that the question of gender and the specifics of female work experiences and their specific positionality in terms of labour have been under-researched and need a stronger focus. Consequently, scholars argue that sectoral specifics need to be considered (Grau-Sarabia and Fuster-Morell, 2021; Kampouri, 2022). The care industry differs fundamentally from the logistics industries where many platform models in food delivery and ride-hailing originate from. This is due to the fact that care functions very differently and on different premises: most notably, a largely female, underpaid and informal workforce that never had the same security through employment as workers from other industries (Kluzik, 2022). Thus, conceptional and empirical papers on care and domestic labour platforms have always aimed to challenge common narratives of labour platforms and platform studies by including invisible and reproductive labour (Ticona and Mateescu, 2018). Consequently, care platform studies aim to include gender and gender-specific tasks into the platform studies queries.
Studies on care platforms often share a similar research design but describe quite different research objects and subjects, very different task profiles, labour forces and operational models (Cornet et al., 2022; Orth and Baum, 2024; Rodríguez-Modroño et al., 2024; Schor et al., 2020). Commonly, however, it is understood as on-site work that takes place in private households, that is, the private realm of care-receivers (Oechselen and Stingl, 2024). This adds to the complexity and power imbalances. In the Indian and South African context, domestic work is more prevalent and became a research focus. The authors show for the domestic work sector in India that intermediation of platforms tends to function parallel to traditional agencies (Tandon and Rathi, 2021, 2023). In the US context, childcare workers and nannies have been a focus, as childcare is not substituted and extremely expensive (Ticona and Mateescu, 2018). In Europe, cleaning workers are most prevalent in studies (e.g. Gerold et al., 2022; Niebler and Animento, 2023; Van Doorn, 2023). As a result, the understanding of what a care platform defines hinges on a broad spectrum of tasks and understandings of care and domestic labour, as well as very different business models from digital mediated care to marketplace platforms. Definitions of care platform labor can be as broad as to include any kind of contracts, platform models and tasks as long as transactions are domestic or household-related (Ustek-Spilda et al., 2022), and as close as to include only self-employed workers (Baum et al., 2020; Baum and Kufner, 2021; Hopwood et al., 2024) and distinctively feature platforms with a ‘focus on actual care in contrast to domestic services (e.g. cleaning, repair works)’ (Trojansky, 2020; similarly, McDonald et al., 2021).
Yet, a rather large set of scholars has conducted research on care platforms through a lens of ‘decent work’ (e.g. Khan et al., 2023; Macdonald, 2021; Pulignano et al., 2023; Vogel and Köszegi, 2024). Within this framework, it is often assumed that domestic and care platforms further undermine working conditions through digital intermediation. Namely, scholars cite misclassification of dependent work as self-employed and platforms’ negation of employer status as the reasons for enhanced precarity. Others have argued that platforms could provide a route to more formalization and positive visibility in formerly informalized professions such as care, where labour has always been informal and precarious (Rani et al., 2022; for overview: Surie and Huws, 2023; Vogel and Köszegi, 2024). However, reoccurring findings cast doubt on the formalization aspect of platforms and argue that formalization remains an imaginary of platforms used to sell services (Mateescu and Ticona, 2020; Pulignano et al., 2023; Ticona and Mateescu, 2018).
Across Europe, evidence show that especially cleaning platforms cater to and exploit specific migrant communities, that is, workers who need to gain income on arrival because they face substantial barriers elsewhere (Floros and Jørgensen, 2022; Niebler and Animento, 2023; Orth, 2024; Van Doorn, 2023). For example, studies found empirical evidence that cleaning labour platforms pay little attention to the documentation provided and allow registration with minimal documentation (Floros and Jørgensen, 2022; Van Doorn, 2023). Yet, as early studies on care pointed out, within the realm of care, this has to be contextualized with the conditions of the care sector and its general workforce being hugely migrant, informal and over-exploited (Hunt and Machingura, 2016). Thus far, the question of what disproportion means in eldercare remains largely unanswered. Nevertheless, the literature provides us with strong arguments of taking the context of the sectors and subsectors into consideration, especially when it comes to migrant work, control and precarization (Glaser, 2021; Khan et al., 2023; Macdonald, 2021).
In sum, the studies on cleaning, domestic and care platforms still feature a broad continuum of relational and household tasks. Nevertheless, the burgeoning scientific debate on care platform labour has thus far mainly provided empirical knowledge about the working conditions of cleaning workers (e.g. Floros, 2024; Gerold et al., 2022; Gruszka and Böhm, 2020; Niebler and Animento, 2023), domestic workers (Hunt and Machingura, 2016; Pereyra et al., 2022; Pulignano et al., 2023; Rani et al., 2022; Tandon and Rathi, 2021, 2023) and childcare workers (e.g. Gebrial and Bettington, 2022; Kalemba et al., 2023; Ticona and Mateescu, 2018; Vogel and Köszegi, 2024), showcasing extremely informal and often precarious working conditions.
The studies mentioned provide a valuable resource for this paper, especially by providing accounts of female labour in platform studies and the overall specifics of care in platform settings that need accounting for. However, there seems to be more variables in that equation. Research on care platforms often feature only very few eldercare workers or assistants to adults with disabilities (e.g. Khan et al., 2023; McDonald et al., 2023; Pulignano et al., 2023). It also remains questionable, whether findings from one care sector (e.g. childcare) can be transferred to others (like elder care) as welfare state regulations can be quite different. Existing empirical evidence seems to suggest that there is a different story to be told for eldercare work and platforms. More recently, scholars have also argued that we need to differentiate between platform types and show how different models function on different grounds and workforce (Hopwood et al., 2024; Orth and Baum, 2024; Rodríguez-Modroño et al., 2024; Tandon and Rathi, 2021, 2023).
Platform Specifics Related to the Field of Eldercare
A small body of empirical research on highly relational care work platforms has evolved. These studies try to tie together the challenges identified for platform work and the challenges identified for care work in the context of crisis narratives (Khan et al., 2023; Macdonald, 2021; Williams et al., 2021). Also, they try to pinpoint the differences to other platform sectors and care fields. For instance, authors showed that workers take negotiations off platforms and display care ethical motivations (Khan et al., 2023). Molitor et al. (2021) show that care workers share only very few features with the platformized workforce in other sectors and that care workers report to be relatively happy with their jobs. However, many have argued that the working conditions decrease also for eldercare workers by application of on-demand and review logics (Pulignano et al., 2023). More care-centred empirical research has challenged the notion of worsening conditions by focusing on one specific type of work, for example, assistant and home care workers remaining closer to the care worker struggles and thus reporting lesser degrees of distress by platform logics (Baum et al., 2020; Glaser, 2021) and also lesser degrees of dependence on platforms (Rodríguez-Modroño et al., 2024).
Research building on multi-task care platforms such as Care.com emphasized the necessity of a specific skill set, especially language and digital skills due to the profile-based set-up of domestic care platforms (Fetterolf, 2022; Orth, 2024; Ticona and Mateescu, 2018; Williams et al., 2021). Moreover, studies from researchers on Australia applied HR perspectives on eldercare and adult care assistance via platforms and highlight that the registration and activity on a domestic care platform resembles the labour process of cloud work graphic design platforms (McDonald et al., 2021; Williams et al., 2021). Arguably, care workers have to put way more effort and invisible work into profiles than riders and delivery drivers, such as advertising skills and profile work (McDonald et al., 2021; Pulignano et al., 2023; Williams et al., 2021). Also, while transport and food delivery platform work does not require formal qualifications, eldercare work is a context where ‘workers are required to possess skills that would usually require a vocational certificate/qualification to obtain employment in a traditional organization’ (Williams et al., 2021: 4143). What is more, these authors also established that the vetting and registration for relational care work platforms has high bars: care workers have to provide proof of citizenship, criminal record cheques and demonstrations of experience (McDonald et al., 2021: 882). Nevertheless, it remains an often-referenced narrative, that workers who use platforms resort to them, due to a lack of skills, and have easy access (Rodríguez-Modroño et al., 2023, 2024). Consequently, this paper takes these findings up and asks for qualification, experience and skill sets of workers. Also, it investigates workers’ care ethical motivations and orientations.
More recently, for the case of Spain, Rodríguez-Modroño et al. (2023, 2024) have built on Tandon and Rathi’s claim that platforms for domestic workers only complement traditional agencies. Based on interview and theoretical sampling, Rodríguez-Modroño et al. investigated into the workforce in digital care intermediation. They argue that different types of workers access different forms of intermediation. However, the case of Spain is specific, because care worker migration follows colonial paths and workers might not face the same language barriers as in other EU countries. Thus, within a German context, it might be a different segmentation at hand.
Overall, the literature on care platforms has made important contributions to our understanding of the business models and on marketplace platforms in general. On-demand models are thus far researched and most prevalent in cleaning, where migrant work seems to dominate. For eldercare, however, most scholars managed to recruit workers dominantly via the so-called multi-task marketplace platforms. These platforms feature a multitude of tasks and only some eldercare workers (Orth and Baum, 2024). Some multi-task platforms sometimes collaborate with researchers, that is, give access to workers in one way or the other (Khan et al., 2023; McDonald et al., 2021; Molitor et al., 2021; Williams et al., 2021). Thus, a few important questions remain: Do workers on on-demand platforms for eldercare mirror the findings form these studies? Do care workers on platforms mirror the off-platform eldercare workforce? Are migrants overrepresented as they are in other forms of platform labour and care labour? And, what are the motivations for workers to do these jobs despite the widely decried bad working conditions and worker scarcity? In short: Who actually works on platforms facilitating eldercare and why? These are the questions this study seeks to answer, taking Germany as an empirical case that allows us to understand this in the context of a welfare state with a specific care regulation.
Care Platforms and the Welfare State: The Context of Marketization and Cash-for-Care Policies
Research has pointed out that for eldercare and care platforms, the welfare state regulation plays a crucial role. For example, eldercare platforms especially thrive, when welfare states introduce cash-for-care policies (Macdonald, 2021; McDonald et al., 2021; Williams et al., 2021). The authors thus argue that these policies foster marketization, commodification, and personalization as modes to make the welfare state more efficient and less expensive (Arcidiacono et al., 2022; Huws, 2019; Williams et al., 2021). Moreover, cash-for-care policies fuel a demand of marketized care services (Macdonald, 2021; Pulignano et al., 2023; Williams et al., 2021). What is more, advancing marketization is especially problematic due to the special nature of care work: Scholars argue that market logics foster the dehumanization of genuine human tasks by making them a clickable commodity and worsen the conditions of care takers evermore (Dowling, 2022; Strüver, 2021).
Nevertheless, within the EU, cash-for-care is granted for the support of activities of the daily life (Blanchard, 2021; Trojansky, 2020). It has become the dominant strategy to fix existing care voids (Hoff and Pottharst, 2023). In the EU today, elderly have a right to access and quality of LTC as part of the Principles in the European Pillar of Social Rights (EPSR). Thus, paid daily assistance for daily support can range from shopping and comfort time spent together to dementia-sensible training, medical and welfare benefits requirements advice. This is to a vast extend a result and part of a marketized care field, where labour is highly regulated, demand is high and care gaps expand in the realm of home care arrangements (Hoff and Pottharst, 2023; Rodríguez-Modroño et al., 2023, 2024; Trojansky, 2020).
Within the German context, additional cash-for-care policies were introduced as part of a constant reconstruction of the care system. While there has been a general care insurance introduced in the 1990s, it is constructed as a market-based system with ageing in place (Hoff and Pottharst, 2023). Thus, 80% of elders with care allowances receive care at home, primarily by relatives (DeStatis, 2022). Although funding is available for daily assistance, a limited number of eligible care recipients take advantage of these allowances (Scheerbaum et al., 2024: 14). Labour in this field has been described as precarious all along, and marketization has increased the pressure on workers (Dowling, 2022). Multiple job holding is not only very common but has increased recently (Schürmann and Gather, 2018).
In this context, the German case is specifically intriguing: The mentioned care gaps are clearly visible and thus different types of platforms operate and offer their fix: In the last 5 years, up to five on-demand platforms solely focused on eldercare in the home. What makes the German case complex is its federal structure and the need for a formal qualification in care. Daily assistance and its reimbursement are regulated on communal levels, which means that there can be very different rules in each regional state for qualification, reimbursements of care services and self-employment. Care platforms that operate in Germany today focus on daily assistance care which can be compensated and reimbursed by small monthly allowances (125 Euros a month, up to 3000 Euro a year depending on your spending on other assistance and severance of care requirements and allowance granted (Baga et al., 2020; Baum, in press; Hoff and Pottharst, 2023)).
In sum, studies including and focusing on domestic and eldercare platforms have in common that intermediate agents have been always part and remain to be part of the care ecosystem which seems connected to how care work is organized (Baum et al., 2020; Blanchard, 2021; Rodríguez-Modroño et al., 2024). Also, the care ecosystem is highly influenced by marketization (Macdonald, 2021) and other sector specifications, like the mainly female workforce that works predominantly part-time. Building on the cited studies, it, nevertheless, often remains a puzzle, how many of the interviewees are eldercare workers, and often the answer is very few (Orth and Baum, 2024). This puzzle sometimes stems from research designing and sampling choices (for overview, Orth and Baum, 2024), but it also seems tremendously difficult to find eldercare workers using platforms. Researchers report huge conundrums finding those workers (Hopwood et al., 2024; Kampouri, 2022; Orth and Baum, 2024). Against this backdrop, the German case is so compelling as it is possible to research care platforms from the angle of workers who choose self-employment deliberately and sample for different platform types.
In addition, the methods applied allow to follow workers motivations, orientations and meaning-making. Within the limitations of the specifics outlaid for the German case, the study can answer the question who actually conducts care work on on-demand eldercare platforms and what differences can be found in comparison with other business models and segments of care.
Research Design and Methods
In light of this, the study builds on a mixed-method approach and is focused on finding workers outside of established paths, targeting eldercare workers on on-demand platforms and recruiting eldercare workers only. The empirical insights from this case draw on two data types: a data set of web-crawled profile data (n = 2144) from a German on-demand care platform and interviews with self-employed care workers (n = 14).
The researched platform offers to match self-employed care workers with care-receivers. For this service, they request a commission of 20%–30%. Self-employment and the according paperwork are mandatory and must be uploaded (ID, self-employment registration).
The number of profiles available from the platform not only allows a quantitative view of who offers care work via platforms but also allows for a qualitative exploration of how care workers advertise and represent themselves through their platform profiles, offering insights into their motivations and orientations for doing this work. Looking into specifics of how platforms organize their profile data, we were able to extract a data set of profiles from one on-demand labour platform by expanding the search parameters on a browser level, storing the data with the programme tool cUrl (Lee, 2024). Accordingly, the platform in this sample features workers who offer assistance to the elderly ranging from household tasks and shopping to basic personal care. Consequently, this data set allows to inquire into a workforce, that research so far has little knowledge of.
To process the web-crawled data, I used insights from Lee (2024) on how to explore data sets with Python and received help from professional coders. The data set includes variables for each worker, such as the hourly pay rate workers ask for, their age, their postcode, their formal qualifications, as well as language skills. In addition, their motivation and experience as texts (string variables).
The self-descriptions workers portray on their profiles were coded manually with an inductive code scheme themed around motivation/selling arguments and experiences (using qualitative text analysis and MaxQDA). To inquire into migration experiences, I ascribed a proxy to identify possible migrant labour on the platform. Also, self-identifying migrants were coded as such. All profile text data were first coded in vivo, and for pre-defined themes, then I established more abstract categories. All manually qualitatively coded variables were then re-integrated as categorial variables into the data set for the explorative descriptive analysis and finally a cluster analysis.
The profile texts must be understood as aspirational. They are written to sell care and might cater to the platform’s affordances and what clients want to hear and might resemble each other more as workers might align their profile with other profiles. However, on this platform and in contrast to others researched, no default text is available. Nevertheless, the profile texts are obviously part of the unpaid and emotional labour that is mandatory for platform workers (Marà and Pulignano, 2022; Pulignano et al., 2023). However, the profile data set shows how workers legitimize their advertising and how and why they commercialize their caretaking, that is, they give an insight into why workers offer care via a platform online, their sense-making and orientations.
The text-based variables added to the data set are grouped by motivation and experience. The motivation was grouped into social/interactional (e.g. ‘Helping makes me feel good’) and value-based motivation (e.g. ‘everybody needs care sometimes’), critical care void (‘elderly people suffer’, ‘care is not good enough’) and passion (e.g. ‘I love my job’, ‘care is my passion’). The experience was grouped into occupational (e.g. eldercare residences or hospitals), occupational support work (e.g. cleaning in hospitals), informal (e.g. paid, in-home care), unpaid (e.g. care for relatives or volunteering) and off-topic (e.g. being a mum or a chef).
To find commonalities among workers, a cluster analysis was deployed as this method can identify underlying groups within large data sets. We used the k-prototypes algorithm according to Huang (1998; γ = 0.5). Based on the k-means and k-modes algorithms, this method is especially suited for real-life data sets including different types of variables (numeric and categorical). The resulting clusters were relatively stable, even when we adjusted the experiences and motivations from the first to the final model and experimented with different numbers of clusters (k). Finally, we used five clusters (k = 5), due to the elbow curve metrics (utilizing the within cluster sum of squares). The Adjusted Rand Index, which was used to measure the stability of the cluster assignments, remained above 0.8 for the model.
The mixed-method design for the study follows a model of conversion mixed design, that is, two empirical strands are done in parallel, but the analysis is combined (Schoonenboom and Johnson, 2017; Tashakkori et al., 2021). Accordingly, I recruited self-employed care workers via different social media channels and directly writing to workers, who self-advertise online and are often affiliated with other agencies (referred to as franchises) and training facilitators. This resulted in 14 in-depth interviews with workers with online appearances, who provide on-site care. For the qualitative data analysis of these interviews, I used an approach of empirical subjectivation research (Traue and Pfahl, 2022) to reconstruct workers’ sense-making and orientations. The introductory narratives of the workers were interpreted in group sessions sequentially and word-for-word. However, this paper prioritizes the data set analysis; the workers’ narratives supplement this analysis for a better understanding of two clusters.
This empirical contribution aims to enhance the discussion of care platforms by allowing insights that were hitherto not possible by focusing on on-demand platform labour and combining quantitative and qualitative data. Combining the two data sets, as well as qualitative and quantitative methods to complement each other, allows insights into the actual workforce behind platforms and the orientations of care workers.
Sample Demographics
The data set sample consists of 2144 worker profiles. Of these workers, 70% are female and 30% are male, seven workers refused to fit themselves into the binary classification. The mean age of the workers is 40 years. In total, 36% of the registered workers have stated to have a qualification in ‘healthcare’ and specified according to the options provided by the platform interface.
The interview sample consists of 15 care workers between 18 and 80 (14 interviews, 1 interview featured two care workers who founded a business together). Six of them are professionally trained care workers, one is a trained medical technician and eight persons had other former professions. Most are German citizens or dual nationals, two are non-German citizens. Four persons identified as male. Three of the total 15 are associated with a ‘care franchise’; 7 of them hold a care certification for self-employed care assistance by a training facilitator. All of them offer their care support services online in one way or the other.
Empirical Findings
The empirical findings will present an in-depth inquiry into the question who actually offers care via on-demand eldercare platforms, mainly based on the explorative data analysis of the data set. A complementary mixed-method approach including a cluster analysis using k-prototypes was applied to the profile data set. In this section, I will first introduce the five clusters. The clusters and the data set are subsequently analysed along the following dimensions: (1) general demographics, including age, gender, regional origin, potential migration status; (2) qualification/occupational groups and experience; (3) rates (wages) and working hours/availability; (4) offered tasks; and (5) motivations and orientations. In the last section, I will complement these findings with the interview data.
Clusters
The cluster analysis provides insight into the types of workers one can find on on-demand eldercare platforms. The cluster analysis with five clusters proved to have reasonable stability. The clusters group around the given dimensions. Variables with the most influence are age, asking rate, and working time, but also regional origin, task, motivation, and experience come into play by applying the k-prototypes algorithm. The five clusters include (1) occupational care workers, (2) mature career changers, (3) young urban helpers, (4) urban high-intensity workers with a higher migrant share, and (5) rural workers with a higher male share. These clusters or segments of workers share the following characteristics. 2
The occupational care workers share a high asking rate (mean: €30.75). Also, two out of three workers in the cluster offer services for more than 10 hours a week. However, the cluster seems not determined by a certain age or gender. Rather, we find workers of all ages in this cluster (mean: 45 years). Compared with the other clusters, this group has the highest share of formally qualified workers (two out of three) and of occupational experience (every second person). They are more likely to offer personal care tasks, but less likely than others to offer housekeeping. They score highest for the stated motivation ‘caring out of passion’ and identification of a ‘critical care void’ they want to attend to.
The mature career changers have a higher share of female workers (80%) than the other cluster and the whole data set. These workers’ asking rate ranges between €15 and €30 per hour. However, compared to the occupational cluster, the average rate is much lower €16.76 (mean). They offer their service for fewer hours; their availability is mostly 5–20 hours per week. One defining factor for the cluster is their age: they tend to be between 45 and 70 years old (mean: 56 years). Also, they have the most practical unpaid experience, but the lowest occupational experience and a lower share of formal qualification. They do not offer personal care tasks, but neither offer housekeeping. This group scores highest for the motivation ‘critical care void’.
The urban youngsters are basically the opposite of the first two groups: They ask for the lowest rates (mean: €15.20) and are comparatively young, that is, under 45 years, with an average age of 31 years (mean). Typically, they are situated in large metropolitan, rather than rural areas. This is the group with the least formal qualifications. However, some social workers are present in this group. The workers of this cluster offer mainly basic supervision of the elderly and have the highest rejection rates for all other tasks. They tend to have a higher score on ‘social motivation’.
The urban (migrant) high-intensity workers ask for relatively low pay (mean: €15.71), but work on average the highest number of hours: 90% of this group offer availability for more than 21 hours. Thus, they shoulder the highest workload. This cluster has the highest migrant share, and workers also originate from cities rather than rural areas. Their age is well distributed. However, there is a decrease after 45 years. This group offers all possible tasks and does not tend to exclude anything. This group has the highest share of ‘human virtue’ and ‘social’ motivations (i.e. they most often state, ‘we should care for each other’ or ‘I want to help people’).
The rural workers are not defined by age or working time but by region, pay and gender. This group has the highest male share (every second person), and the workers stem mainly from rural areas. Their pay is the second highest (mean: €18.24). The cluster features a high share of rural formally qualified workers, and every second person also reports occupational experience, also from occupations such as physiotherapists. The cluster is also characterized by offering more tech support than all others, but housekeeping and shopping support are also in their portfolios. They do not spark out in terms of motivation.
The cluster analysis shows a variety of worker types on the researched on-demand platform in terms of demographics, pay and working hours, qualification and experience, offered services and motivations. These dimensions stem from a combination of qualitative coding the texts for motivation, experience and potential migration status, as well as categorical and metrical variables from the data set (e.g. asking rate, working hours, qualification). The following section will give an overview of the variety within each dimension.
Demographics: Age, Gender, Region, Migration Status
On the platform studied, the majority, that is, 95%, of all profile owners are between 20 and 60 years old; the median age is 40 years. The majority (70%) of workers self-identify as female, which is in accordance with the overall occupational gender division within the care sector. Only 20% of the workers were classified as potentially being migrants. Notably, only every second person lives in an urban setting. The data illustrate that the on-demand care workforce on this platform is mainly female, workers are in their active working age, live in both urban and rural areas, and migrants are not overrepresented.
Moreover, the sample also shows that there are no certain migrant groups that stand out. More concretely, having used language skills, text impression a potential migrant status. This does not indicate any overrepresentation. Rather, when language skills are taken as a proxy (i.e. Russian, Spanish, Turkish, Polish, Arabic, Italy and former Yugoslavia), the potential migrant backgrounds on this on-demand platform match well with migration patterns of long-settled migrants in Germany.
These quantitative findings offer a stark contrast to other platform labour studies, in which samples mainly consist of male and urban workers (e.g. Beckmann et al., 2024) and recent arrivals in the city (e.g. Van Doorn, 2023). Similarly, the data challenge narratives of care platform representatives who often claim their workforce primarily comprises students and pensioners (e.g. Kasten, 2022). Instead, the data for the studied platform suggest that it is mainly well-settled workers in their active working age who engage with care platforms.
Qualification, Occupational Groups and Experience
As platform labour research often portrays this labour as low-skilled, easily accessible, and a work opportunity of last resort, a closer look at the qualifications and experience of workers is particularly interesting. Compared with this notion, the profiles generallly reveal that workers have substantial prior experience in caretaking.
First, one of three workers on the platform is formally qualified in the care sector, for example, has a vocational qualification in nursing (36% of workers in the data). Most common are care workers with assistance qualifications, that is, they are trained (2 years) for assistance in nursing and eldercare. Accordingly, 74% of the workers in the data set claim no formal qualification. However, second, the analysis showed that they still describe a wealth of experience. These include not only informal but paid eldercare, occupational experience in a social profession other than eldercare but also more informal care experiences, such as unpaid care for relatives, internships in care homes, hospitals, and assistance to people with disabilities. What is more, profiles also reference years of voluntary practice and support of the elderly. At the least, profiles report experiences as mothers, fathers, brothers and grandchildren. Often, workers provide a combination of caring for disabled, young and elder family members, volunteering/interning and even certification. As a result, only very few workers (10%) state to have no experience at all. Overall, even when workers consider themselves not formally qualified, their profiles portray extensive experiences in caretaking.
Rates (Wages) and Working Time
For the whole sample, the average hourly wage workers ask for is €17.63. However, formal qualification makes a huge difference for pay, as is also evident in the clusters: Workers with a formal qualification earn almost €4 (€20.74) more than those without a formal qualification (€16.20). Moreover, care workers from rural areas also ask for significantly higher rates than urban workers. However, what seems striking is the fact that formally qualified care workers can be found in all clusters, including the low-paid clusters. As the cluster analysis shows, the background of the workers, that is age, experience, rural/urban, also highly influences their asking rates.
The overall majority of workers ask for rates between €12.50 and €30.00. Interestingly, the men in the sample ask for significantly higher rates (€18.25 vs €17.36). Notably, the gender pay between men and women in the segment of workers without formal qualifications is €1.21 per hour. This does not substantially change with higher qualifications. Only in the lowest qualification class formed, the pay is almost equal. However, there are only very few men with this qualification. Notably, for the vocational-trained assistants and the professionally trained care workers, a severe gender pay gap remains: qualified men earn €1.38 more than equally qualified women (€21.83 vs €19.20).
With regard to how many hours people work, the data set indicates most workers are available on a part-time basis. Half of the workers are only available for up to 10 hours a week. The other half is evenly distributed between 11 and 40 hours. Men not only have higher asking rates, but they also tend to offer higher availability: Men offer an average of 15 hours, whereas the average for women is only 10 hours. Interestingly, workers, and this is especially poignant in the segment of the formally qualified workers, report in their profiles to also work jobs in eldercare homes and thus offer some additional self-employed care work on the side.
Tasks
The care workers can choose between seven different tasks they can offer as a service to clients. Care is divided into tasks such as supervision, guidance, mobile assistance, housekeeping, shopping, technical support and basic personal and physical care. Most of these tasks do not require a formal qualification. The German healthcare system has strict legal regulations and professional standards that cover most care tasks. Therefore, non-formally qualified workers can legally only offer basic personal care tasks. In addition, the reimbursement rules for daily assistance seem to define the tasks available. The vast majority of workers in the data set offer five to seven different services. However, only every third person offers basic personal and physical care services. The same applies to technical support. The most prevalent (80%–90%) is care in the sense of supervision and guidance. Nevertheless, there is also a latent division of the sample by the tasks offered. The clusters show a specific task set, the more professional the cluster presents, the less housekeeping and shopping support is on offer. The rather male and rural cluster seems characterized by its overrepresentation of tech support which is much lower for all other groups. The urban helpers mainly offer basic supervision. The (more migrant) urban high-intensity care workers basically offer all tasks on hand.
Motivations and Orientations
Overall, care workers on platforms portray themselves as being passionate about care. Even when commodified and sold via a platform, care is portrayed as a service out of love and grounded in virtue and care ethics. The reasoning of the care workers has been coded into four groups: social motivation, virtue/humanistic, critical care void and passion for care. The social and humanistic reasoning categories are highly prevalent across all segments in the data set: every second person references such an orientation and reasoning. Workers argue that they offer care because they like to work with humans/elderly, appreciate a smile and specific gratitude received in return (social motivation) and state – almost equally distributed among all clusters – to hope for reciprocity, acknowledge the right to quality of care for everyone and believe that everyone should live in dignity and it is a virtue to care for others (virtue-based/humanistic motivation).
Especially, formally high-qualified care workers tend to state that they love their job and regard it as their passion and as a life calling. Thus, they provide care ‘out of passion’. They also mention more often than others that offering their services on the platform in addition to their occupation, as it means less time constraints. Given their passion for their jobs, they also reason with a critical care void. Also, the informally experienced mature career changers tend to cite their own experience with a critical care void as a reason for having entered the care vocation.
In sum, the data set analyis shows a very qualified, experienced and highly socially motivated workforce. This stands out in comparison with the narrative of unskilled work on care platforms. In line with previous care platform studies and the segmentation of care work within society at large, the workers on this on-demand platform tend to be largely female. However, the workers are neither especially young nor old, but in their active working age. Also, no overrepresentation of migrant workers or newly arrived migrants could be found. However, the cluster analysis provides insight into the diverse groups that can be found on on-demand care platforms. Their different backgrounds, especially urban/rural, male/female, younger/older and the prevalence of occupational or informal experience, grealty influence the stratification of how much pay they ask for.
Combining the Data Set with Interview Data: Paths into Commodified Care
To complement these findings from the profile descriptions of care workers registered with a German on-demand care platform, I also conducted 14 worker interviews with self-employed care workers. The aim of these narrative interviews was to better understand workers’ motivations, orientations and reasons for self-employment, and explore why they decide to offer their care via a platform or not. What is more, the interviews dive deeper into two central themes that seem to connect all workers and could be found in the orientations of the workers’ profiles.
The first finding from the profile analysis was that people start offering care services online based on former contact points, that is, experience and a practice of caretaking, either occupational or within the family or neighbourly settings. Second, there seems to be a common sense of purpose and care ethics driving the provision of care to others. In the profile texts, workers strive for reciprocity, dignity, and helping as a societal and human virtue. Workers reference a sense of how it feels to be alone with caretaking, and bad experiences with neglect of the elderly in occupational settings in their profile texts. Both themes also emerged strongly from the interviews and were even more salient in interviewee’s accounts of their professional lives.
In addition, interviewees provided longer narratives of how they became self-employed in care and why, showing both tendencies, a social motivation and a surge for meaningful work aligning to their values as well as impulses by a care situation or occupational experiences. The interviewees show a similar division as the data set: one out of three workers has formal care qualification and even more referenced occupational experiences.
Workers formerly employed as professionals commonly decry caretaking in institutions structured by for-profit logics and rigid time frames. These are at odds with their wishes for a work–life balance and meaningful caretaking and cannot be combined with their family obligations. They deliberately chose self-employment, either with a franchise that helps them with welfare state regulation and reimbursement or utilizing their professional status to found a small company for caretaking without additional help: Doing so, most have come across platforms in their online reserach as options but deliberately did not choose this path.
The other interviewees professionalized into self-employed caretakers for two reasons: First, they describe their journey from caretaking for a relative and/or child, resulting in reduced options to return to their professions (typically a female narrative). In this void of opportunities, care as a vocation then becomes a reasonable alternative. Second, they describe a professional reorientation phase, deliberately searching for a more meaningful job with the situation they faced in mind: As Christoph and Maren, two interviewees state, ‘I realized there is a huge demand’, ‘guidance/daily assistance was just not available back then’. Thus, they professionalized their caretaking via short vocational trainings, referencing those certifications and emphasizing the need for professional daily assistance. This approach becomes especially meaningful when they already gained welfare state expertise, that is, knowing how to deal with care allowance, or they are organizationally adept for their former business or managerial careers. Most of the interviewed workers fit very well with the career change cluster, regarding age, motivation, and tasks. They often describe doing ‘zero care and zero housekeeping’, as Elisa, one interviewee put it. This specific understanding of care support could be also identified as a feature of the cluster. Moreover, workers tend to portray themselves as a substantial hinge between medical care and housekeeping. Interviewees often self-identify as organizers of care that substitute relatives. In contrast to the career change clusters, however, the interviewed careworkers asked for higher rates (€30–€40).
A theme to thread all these workers together is that care workers aim to make their gained expertise available to others. Often, they learned how the care and welfare system works, either from occupation or from personal experience and additional trainings. Also, within the data set 10%–15% can be categorized as similarly new home care professionals. Daily assistance (including application for care allowances) often remains the major care gap in the current organization of eldercare in Germany. In interviews and in profile texts alike, workers and especially new professionals portray and advertise a service attitude that caters to this void and is very care-receiver-centred. They reference care ethical stances, including having enough time or acknowledging the right to stay independent and at home as long as possible: in order to achieve this, this type of workers offer substantial support and expertise. They are experts on applying for care allowances that are underutilized and help manoeuvre evolving care needs. The interviews narratives show, in addition to what the profile texts transport, that workers first do short vocational trainings after facing challenging personal care experience. This impacts their orientations. In order to find clients, some also registered on platforms in the aftermath. Mainly, the interviewees possess and offer expertise for dealing with welfare state requirements and emphasize that their services can be directly billed to the care insurances, as they cater to the specific needs and care gaps of the elderly at home with care allowances.
However, finding clients was not a problem for any of the interviewed care providers, and most preferred working self-employed: save one, none even wanted to imagine to change to occupational employed care. While most interviewed workers were critical of platform models, they also either had several dead profiles as a leftover of the time when they started or they relied mostly on their franchise or training facilitators for online representation. Nevertheless, all workers agreed that the best way to find clients was in real life and preferred analogue marketing. Regardless of their utilization, however, most interviewed workers were critical of platforms, as for them, platforms also represent a marketization and for-profit logic. In opposition, they perceive themselves as directly paid and thus less commodified. Also, they strongly advocate for more visibility and appreciation of informal care work in society. Thus, save one, most did neither describe their work as commodified nor identified as precarious.
Despite these differences, in the German home care society where care labour is chronically short, interviewee accounts and profiles alike feature examples from their own care experience of a critical void. Workers’ narratives stress how severe this is. Conversely, this also means that workers seem not very dependent on platform work, least the high-earning workers.
In sum, the interviewees not only show a similar orientation but also highlight how workers professionalize into caretaking. However, in their own perception, this work is less commodified than other options of occupational care work. Their orientation is equally and substantially socially motivated and virtue-grounded.
Discussion
The empirical findings show a variety of worker types can be found on the researched on-demand platform for eldercare. The tasks offered are in line with the European system of daily assistance for caretaking in the home, but workers cater to them differently. For some, working self-employed in care and/or offering care services online and on care platforms can become a reasonable additional source of income and meaning (occupational care workers). The interviewees give an impression of why workers are motivated to enter care laterally and become home care professionals. Their narratives match with the motivations and orientations found in profile texts (career changers). In contrast to findings from other sectors (Floros and Jørgensen, 2022; Van Doorn, 2023), being registered with an on-demand care work seems much more grounded in qualification and experience in eldercare. This has been already pointed out by research on marketplace platforms in care (McDonald et al., 2021, 2023).
Furthermore, the data show the importance of informal and unpaid experience for platform care work. This corroborates findings from recent papers that have argued a stratification and segmentation among care workers on different types of agents (Rodríguez-Modroño et al., 2024). However, the presented findings show that for on-demand care work, formally qualified workers can be found alongside those who just want to help out based on virtue and social motivations. While studies commonly problematize this as competition that devalues care work (Dowling, 2022), the findings suggest workers can and do charge high rates based on their qualifications nevertheless. Strikingly, men tend to better utilize commercialized care, though the unpaid experience of women seems much higher. A unique contribution of this research is to prove a gender pay gap for on-demand platforms in eldercare. Men also offer higher availability. However, taking the high-intensity workers and the mature career changers into account, it does not seem to be a linear connection between working hours and pay. The cluster analysis indicates the structure behind this: With a higher share of either male or migrant care workers, the working time/availability tends to be higher, whereas the professional workers and the lateral entries work fewer hours.
What is more, in comparison with the prevalence of migrants among care workers in institutional settings, such as eldercare residencies and hospitals, and in live-in care, migrants seem to be underrepresented in this data set or at least not visible as much. Overall, this seems to support other researchers’ arguments that care work via platforms is not easily accessible and requires certain skill sets (McDonald et al., 2023) and that workers are segmented strongly along specific care tasks (Rodríguez-Modroño et al., 2024).
Pay and Working Conditions Contextualized
The cluster analysis visualizes the different groups and pay variance on platforms. Some workers seem to earn comparably low wages, have little prior experience, but also only offer few hours of basic care, specifically supervision. However, half of the workers on the researched on-demand platform ask for higher rates and offer semi-professional care services grounded in substantial experience. Nevertheless, the pay rate for labour without the EU standard welfare state securities (i.e. unemployment, health, and care insurance) seems initially low.
While the discussion of decent work on platforms argues that rates and conditions are insufficient for generating a substantial monthly income, especially if the employment status is independent and no social security is included (cf. McDonald et al., 2023; Pulignano et al., 2023; Woodcock and Graham, 2020), the asking rates on the platform, nevertheless, match pay in care homes and hospitals for the level of qualification present. At the time of the data collection, the minimum wage for ‘unqualified’ care assistance and support workers in hospitals and in care homes was below €11 and €15 for qualified staff (formal 2–3 years training) before taxes and social security (Federal Department of Health, 2022).
First, wages in the care industry in Germany by and large do not provide a living wage and thus make earning an extra income necessary. Second, within a hybrid work arrangement, where employment in the care industry is combined with other paid care labour, this hybrid arrangement secures that mandatory social security costs are covered by an employer. 3 A phenomenon Schor et al. (2020) call ‘free-riding on security by platforms’. For workers, this means they are not left entirely without a social security net and income from additional platform labour may seem reasonable in order to make ends meet. This seems to be the case for many of the occupational care workers in the data set. Third, care duties within their personal realm may mean for many women a need for flexible earning opportunities, fewer job chances, and less pay. Almost all female workers interviewed reported this struggle and appreciated the autonomy of self-employment for setting their own working times according to their own care obligations and self-care needs. Especially, the former care workers found it unbearable to care at home and occupationally at the same time.
Hence, platforms can be associated with two aspects: first, visualizing the effects of marketization and underpayment in care, and second, free-riding not only on the security of care employments (Schor et al., 2020) but also on the undervalued care experiences and qualifications of workers.
Motives and Orientations: Why Do Workers Care Self-Employed?
The motivations and orientations of care workers included in this study show the implications of marketization of care, but do not show negative more marketized working conditions for care workers who are self-employed and/or use platforms, mirroring recent findings by researchers in the United Kingdom (Muldoon et al., 2024). Rather, the present research showcases that care (and social) workers in institutional settings need extra income, struggle with the conditions in care and thus provide care outside of these settings. Those working professionally, most often frame care as a passion and offer additional support or switch to self-employment to provide better or ‘decent care’. It is crucial to understand their choices for self-employment against the backdrop of under-evaluation and underpayment of care labour under capitalist conditions (Aulenbacher et al., 2018; Fraser, 2016).
Notably, even though workers choose to provide care work as a commodified occupation, that is, self-employed and/or facilitated via an on-demand platform, they do so by deploying care ethical and care-centred motivation. This is prevalent in both data types. Most prevalent in the in-depth interviews with care workers is their self-representation as guides through bureaucracy and challenges people have to go through when an independent life starts to diminish. With the over-exhausted workforce of care workers and doctors in mind, the elderly need evermore guidance and assistance. In contrast, conventional caretaking is portrayed as driven by for-profit logics and rigid time regimes. Professional and semi-professional care workers, that is, those with experience in occupational care and formally qualified, and those who progressed into care as a profession, show great commitment to the cause, and provide support they deem otherwise not available in their reasoning. This is grounded in their experience in caretaking. In addition, people move laterally into the field of home care for finding meaningful work. They are motivated by social and normative values of caretaking and the visible lack thereof. What is more, the interviewed workers really do not (anymore) strive for employment in marketized occupational care settings, as it lacks flexibility and, more importantly, autonomy. Especially, for eldercare workers, care ethics and care ethical considerations largely play into workers’ reasoning, as researchers already suggested (Khan et al., 2023).
Commensurate with findings from Australia that emphasize the role of the welfare state (Macdonald, 2021), all interviewed workers highlight the importance of being someone to guide their clients through cash-for-care bureaucracies. The interviewed care workers present and understand themselves as (semi)professionals who hope to do ‘their part’ by mending the severe symptoms of the care crisis within their direct reach and realm, that is, in their own neighbourhoods. Their ability to address this demand for assistance under care allowance schemes not only offers them both a material opportunity to get reimbursed for their work but also speaks to their values because they can share expert knowledge and help people. However, the interviewees show that platforms are not the route to take for accessing these funds (Baum, 2024, in press). Platforms can be a vehicle to find clients, but many other channels of recruiting are way more fruitful from the perspective of the workers. For others, as the data set exemplifies, it seems a good choice for supplementing income with a hybrid model of caretaking. However, for those who want to professionalize, other agents provide better opportunities for accessing cash-for-care funds.
While platform labour studies tend to centre around questions of fair work and precarity, this analysis challenges some of these assumptions: self-employed care work does not seem to be a last resort, but a choice against marketization and a route to ‘decent care’. Paradoxically, platforms can be a vehicle for this orientation. Within eldercare, it is not the platform model that leads to precarity but rather the under-evaluation of care work at large. As care work tends to be either entirely unpaid or severely underpaid, care workers on platforms as well as self-employed workers off platforms do not describe their work and care arrangement being self-employed as any more precarious than their previous work arrangements. Rather, they utilize their experience in caretaking in order to deal with deeply precarious care arrangements. In their view, it is not their working arrangement that is precarious, but the societal arrangement of caretaking as a whole. In light of precarious working conditions in care, workers want to provide decent care and as first priority. They are willing to pay the costs of self-employment for this goal, as often no better options are available in order to combine care work requirements.
Nevertheless, interviewed workers find platforms just as critical as journalists and researchers. However, some may still use them as a tool to find clients and, most importantly, help out those who face a situation of stark undersupply of care.
Contribution: Who Offers Care Via Platforms?
Building on this, the analysis yields three main findings.
First, the care platform market is a niche market and in general it seems that every care platform tends to build on a very different, niche workforce. In this case, it is well-settled experienced care workers. The findings show that workers offering eldercare online – on platforms or elsewhere – do not seem to have much in common with the workforce described for other platforms in care, especially the most prevalent findings on cleaning and childcare. So even within the sector of care, there are many differences across occupations. Rather, the workforce matches the occupational, socio-demographics and stratifications of the subsector, that is also structured by welfare state regulations. Moreover, the results show that the business model of the platform, that is, marketplace and on-demand, can make a huge difference in the socio-demographic and qualification make-up of the workforce (compare to Molitor et al., 2021).
Second, the data show that care workers on the researched on-demand platform have substantial experience in practical eldercare, including occupational, informal and unpaid caretaking. The in-depth interviews provided additional insights on why different clusters of workers may choose, first, self-employment and, second, may use a platform to find clients. Those who enter caretaking laterally often strive for more meaningful work, autonomy and flexibility. They deliberately choose self-employment as it provides more agency over working conditions and time frames for caretaking: for former care workers, a combination with family and self-care commitments matters as this combination seems impossible to them in marketized care institutions. Care workers want and do utilize their experiences with caretaking and the welfare state organizational requirements. They seek to help others to overcome challenges they once faced themselves or feel adept to overcome due to their care or former managerial occupations. In accordance with the findings from Australia, this shows that the welfare state context matters greatly for eldercare provisions via platforms and other agencies.
Third, in contrast to the findings for childcare and cleaning work intermediated by platforms all over Europe, on the researched platform, no comparable overrepresentation of migrant workers can be found. Rather, against the backdrop that migrant workers take over a vast majority of informal home care work, especially in live-in settings, these workers seem underrepresented. The findings from the data set stand in contrast to assumptions from most care platform studies, that migrant work is disproportionately represented on care platforms in general. However, the presented findings are in accordance with the findings from Molitor et al. (2021) for a marketplace platform, showing that care workers are often female and well-settled workers, and there is no overrepresentation of migrants. Also, the presented findings support and advance the findings of Rodríguez-Modroño et al. (2022, 2023, 2024) that care workers stratify among different platforms and that each care sector and platform model builds on a specific workforce. This ultimately means that it is worthy and necessary to disentangle care and domestic labour and acknowledge that platforms in Europe operate in very specific niches of welfare state gaps.
Limitations and Future Research
Analysing web-crawled data comes with inherent limitations. Whereas most quantitative data analysis on platforms is based on questionnaires, this analysis could only utilize the variables included in the data set. Consequently, it was not possible to check for household income, or the number of hours people report to actually work. However, this limitation was addressed by the mixed-method design that equally considers 14 in-depth interviews and looks into the motivations, orientations, and situatedness of the workers. Nevertheless, the workers interviewed give only additional information for two specific sub-groups of the data set. The workers interviewed work at their full-time capacity and deliberately self-employed and also use other facilitators as the main intermediate. For future research, it seems sensible to find and sample the presented sub-groups defined in the clusters. However, if sampling over platforms is not a good option (Orth and Baum, 2024), this represents a major challenge for research. The findings also provide some insights on why that is: workers may either identify as care workers and not mainly with the way of intermediation (similarly, Tandon and Sekharan, 2022), or in the case of the Urban Helpers, not as workers at all. The recruiting method showed that when targeted directly as self-employed eldercare workers, outcomes improved. In future research, similar data sets could provide more information on factors that affect wages/pay on platforms.
Conclusion
This study has investigated the workforce engaged on a German on-demand platform, uncovering vital insights into the motivations and orientations of care workers who opt for self-employment in the realms of eldercare and home care. These results lend strong support to the assumption that platform labour in eldercare functions very differently to childcare platforms and domestic care platforms.
The data analysis reveals that in self-employed eldercare workers are in majority female, predominantly well-settled, experienced in care giving and of all ages. Remarkably, self-employment – and by extension platform models – may provide a window of opportunity for individuals from (social) care occupations, as well as people with substantial experience in unpaid and informal home care, as well as organizationally adept workers who aspire a career change. Along the dimensions of age, income, qualifications/experience, tasks, and motivations, I identified five distinct clusters of care workers: professional care workers, mature career changers, urban helpers, urban (and migrant) high-intensity workers, and (male) rural workers. This clustering illustrates the diversity of care workers registered with the researched on-demand platform. Despite this heterogeneity, a common thread prevails: workers typically engage in self-employed care following practical experience and first encounters with caretaking and support of elders within their occupational or personal contexts. In addition, workers tend to share high social and value-based motivations. Those who professionalize demonstrate stronger care ethical orientations.
The landscape of home care across Europe, particularly in Germany, is intricately shaped by unpaid caregiving, welfare state regulations, and marketization, resulting in a system characterized by informal and migrant labour. European legislation recognizes a right to LTC and stipulates care quality, which, in conjunction with national policies, intensifies care marketization, namely through privatization and personalization of caretaking. Consequently, home care across Europe and Germany can be subsidized via cash-for-care funding mechanisms. This has a crucial impact on the ecosystem of home care, and ultimately on intermediation. In this context, establishing a business that leverages existing care qualifications and experience has emerged as an alternative and supplementary path to traditional care employment in a market-driven care environment. Notably, workers not only utilize platforms but also collaborate with other agents, franchises and training facilitators to bridge existing caregiving gaps, provide decent care and have a meaninful job.
While platforms feature a diverse array of workers, who do have multiple choices for finding clients, platforms nevertheless represent a free-ride logic. The analysis indicates that platforms free ride not only on the security of occupational care workers staying employed, but also on the workers qualifications and their experiences. However, in comparison with other workers in the care sectors using platforms, the overall dependence of workers on these platforms remains rather low.
The analysis yielded three conclusions: first, care platforms operate in niche markets catering to different workforces; second, workers respond to marketization in care by leveraging their experiences and utilze their qualifications to pursue care self-employment; and third, contrary to trends observed in other areas of platformized care, migrant workers are not disproportionately represented within eldercare platforms in Germany.
Footnotes
Acknowledgements
The coding of the analysis in Python was to a significant extent done by Tina Baumann, who generously coded and explored ways to let the data speak to us. I greatly appreciate the skill, patience and brightness, you put into this project which elevated my research to new levels. My eternal gratitude to the programmer who explored the organization of profile data with me, resulting in the data set explored. Additional thanks to my supervisors and colleagues from Care Transformations Project, namely Tanja Carstensen, Daniela Rastetter, Wolfgang Menz, and Almut Peukert for allocation of funds and the ethical guidance on how to handle and proceed with the web-crawled data. Last but not least, a special thank you and acknowledgement from the bottom of my heart to all the hard-working care workers, those who were so generous to talk to me and trust me with their stories, and also those who were by my side and put me back together: Without (your) care, (this) research would not exist.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: The study and preliminary findings this article builds on started in 2020 under the project Care Transformations at the University of Hamburg (LFF ‘Sorgetransformationen’; BWFG Hamburg federal research funding). Subsequently, it received funding by the University of Hamburg State Graduate Funding Program Scholarships from September 2023 to August 2024 and much appreciated funding by the Equal Opportunity Unit of the University of Hamburg and the Hamburg University Graduate School for transcription of interviews.
