Edited by Christian Suter Jacinto Cuvi Philip Balsiger Mihaela Nedelcu The Future of Work The Future of Work Edited by Christian Suter, Jacinto Cuvi, Philip Balsiger, Mihaela Nedelcu The Future of Work Edited by Christian Suter, Jacinto Cuvi, Philip Balsiger, Mihaela Nedelcu Published with the financial support of the Faculty of Arts and Humanities of the University of Neuchâtel and the Swiss National Science Foundation. Published by Seismo Press, Social sciences and social issues AG, Zurich and Geneva www.seismopress.ch | books@seismopress.ch Seismo Press is supported by the Federal Office of Culture for the years 2021–2024. Text © The authors 2021 ISBN (Print): 978-3-03777-244-7 ISBN (PDF): 978-3-03777-818-0 DOI 10.33058/seismo.30818 Cover: Wessinger und Peng, Zurich and Stuttgart The cover is based on the poster made for the congress “The Future of Work” by Salomé Donzallaz, Neuchâtel This license allows reusers to copy and distribute the material in any medium or format in unadapted form only, for noncommercial purposes only, and only so long as attribution is given to the creator. 5 Content 1 The New World of Work: Current Trends and Uncertain Prospects 7 Jacinto Cuvi, Camille Budon, and Christian Suter Part I Trends and Scenarios of the Future of Work 27 2 Sens et avenir du travail en Europe 29 Dominique Méda 3 Trajectories of Change at Work: A Global Look at 52 Technological Change and Informalization Chris Tilly Part II Digital Transformations 73 4 The Increasing and Invisible Impact of the Working 75 Consumer on Paid Work George Ritzer and Piergiorgio Degli Esposti 5 Varieties of Internet Platforms and their 100 Transformative Capacity Ulrich Dolata Part III Platform Work 117 6 From Commodification to Free Labour: The Gendered 119 Effects of Marketplace Platforms on Work Anne Jourdain 7 Ambivalent Flexibility: Crowd Workers Between 146 Self-Determined Work and Capitalist (Self-)Exploitation on Digital Platforms Eltje Gajewski 6 Part IV Work at the Margins 169 8 The Role of Work Before and After Retirement on 171 Poverty Dynamics in Old Age. Evidence from a Follow-Up Study in Switzerland Rainer Gabriel, Michel Oris, Sonja Kubat, Kushtrim Adili, and Monika Götzö 9 ‘Simple Jobs’ for Disqualified Workers: Employability 199 at the Bottom of the Labour Market Eva Nadai and Anna Gonon 10 Éloignement de l’emploi : trois dimensions du registre 222 de la honte Jean-Pierre Tabin, Monika Piecek, Isabelle Probst et Céline Perrin Part V The Future of Work for Academics 245 11 Partir un jour, sans retour ? Injonctions et pratiques de 247 la mobilité internationale dans l’espace académique suisse contemporain – le cas des postdocs Pierre Bataille et Marie Sautier 12 Career Starts of Swiss Sociologists: Labour Market and 277 Working Conditions for Sociologists Rainer Diaz-Bone About the Authors 299 7 1 The New World of Work: Current Trends and Uncertain Prospects Jacinto Cuvi, Camille Budon, and Christian Suter 1.1 Introduction Is work disappearing or simply changing? If changing, is it set to become increasingly precarious for everyone, across occupations, income brackets, and world regions? And how do those at the margins of the labour market make a living and experience their social exclusion? This volume addresses these and related questions in the wake of the Covid-19 pandemic, at a time of economic crisis and deep uncertainty facing workers around the globe. Contributions were selected from presentations at the Congress of the Swiss Sociological Association on the Future of Work hosted by the University of Neuchatel between September 10 and 12, 2019. They were turned into chapters in the course of 2020, as governments and societies around the world grappled with the pandemic and its economic fallout. The book thus captures the trends that have been reshaping work and labour markets for several decades as well as the impacts of a recent global event. Work and its uncertain future(s) were popular topics of social scientific research and debate long before the pandemic, with echoes reaching far be- yond academic circles. In fact, the changing world of work and the daunting fate of workers is one of those few areas on which the interests of scholars, the media, and the public spontaneously converge. A growing literature examines the general trends of precarization and digitalization as well as more specific issues such as youth unemployment (Weil et al. 2005) or the relations between work and gender (Bettio et al. 2013; Gatta 2014) or work and immigration (Raess and Burgoon 2015). As Simms (2019, 27) notes in What do We Know and What Should We Know about the Future of Work? ‘[t]he current tone of the debate is one of anxiety about whether paid employment will continue in the near future’. 8 There is good reason for such anxiety. The labour market looked grim even before the Covid-19 pandemic. According to Katz and Krueger (2019), precarious work accounts for virtually all net job creation since the 2008 financial crisis in the US, where estimates of the workforce employed in non-standard work arrangements before the pandemic ranged between 10 and 15.8 percent. Likewise, Gutiérrez Barbarrusa (2016) maps the rise in precarious work in Europe between 1995 and 2015 and finds that job in- security affects over one-fourth of the workforce, up from about one fifth before the 2008 financial crisis (see also OCDE 2014). In Switzerland, too, there is a (slight) increase in the proportion of people affected by atypical and precarious labour conditions such as involuntary part-time work, reduced working hours, temporary jobs which amount to about 15 percent of the entire working population in 2014 (Ehrler et al. 2016, 49). While workers’ reasons to engage in part-time work vary and there is much diversity in the profiles of workers who engage in atypical employment (Bühlmann 2013) as well as in their strategies to cope with precarious work (Preoteasa et al. 2016), the rise of part-time is a long-running trend that began in the 1970s (OFS 2006; OFS 2020a). From a theoretical perspective, the spread of non-standard employment relations reflects an expansion of the market mechanism both inside the firm and beyond its boundaries, where self-employed workers compete for gigs. Whereas internal labour markets formerly insulated workers from competition, especially in large organizations, and provided employees with incentives to invest in the growth of their firm, workers increasingly find themselves on the lookout for opportunities everywhere, vying for jobs—including their own—with co-workers and jobseekers outside their organization (Spreitzer et al. 2017). As Cappelli et al. (1997, 15) note, ‘[u]nder the new arrangements, employees share much more in the risks of doing business, take on more of the responsibility for managing their own careers, and find that their relation- ship with management is governed to a greater extent […] by market forces.’ Much of the existing literature on the ongoing transformation of work seeks to map out the landscape and trace its origins. The rise of neoliberalism is a recurrent theme in the literature on precarization (Standing 2011). But studies also point to more specific mechanisms, such as the advent of the financial model of the firm, which made shareholder value maximization the single focus of management (Kalleberg and Vallas 2017). Restructured firms resort to outsourcing and downsizing even during times of financial stability. Other causal factors include the secular decline of unions and labour power as well as globalization, which fosters worldwide competition among 9 workers in a context of growing trade and unhindered capital flows (Silver 2003; Stiglitz 2001; Suter 2009). Another strand of research is centred on prospective analysis and policy implications, tackling such areas as workforce skill development or labour market regulation. A seminal paper by Richard Baldwin (2006) thus notes that new information and communication technologies (ICTs) upset conventional assumptions when planning for the future of work in Europe. Digital ICTs allow for the outsourcing not only of manufacturing, but also services, including high-skill tasks. Hence, a massive educational investment in digital training can backfire in a world where software engineers half-way around the globe can perform the same tasks at a lower pay. Likewise, a report commissioned by the European Parliament recommends ‘anticipatory policies’ to ensure social protection for workers in platform mediated employment, 70 percent of whom reported lacking access to basic welfare schemes (Forde et al. 2017). It is not a coincidence that digital technologies loom large in the poli- cy-oriented literature on the future of work. They hold the promise of drastic increases in productivity and carry the threat of massive layoffs at the same time. In fact, two centuries after the Luddites, technological fear is making a comeback among the public, with worries that computers or self-operating machines might replace humans in most tasks. The bots-eat-jobs scenario looms behind the catchphrase ‘Industry 4.0’, which describes a stage in digi- talized production at which machines become autonomous thanks to machine learning and other forms of artificial intelligence, and at which computers take control of production processes through interconnection with physical equipment in so-called cyber-physical systems (Lasi et al. 2014). It is common practice in academic and private-sector studies of the future of work to make informed predictions about the type and number of jobs that will be lost to automation (Susskind and Susskind 2015; Manyika et al. 2017). The thrust of this volume is to take a more comprehensive view that considers the effects of digitalization not only on the disappearance of paid employment but also on the everyday experience of those who continue to be employed, online and offline, and to flesh out the social context wherein technical change takes place. Given the relative novelty of the platform and gig economies, there is still much that we do not know about the ramifica- tions of platform work on the lives of workers. Contributions to this volume consider trends, like informalization, that affect labour markets around the world (e.g. Tilly, Chapter 3) and they examine the effects of digitalization on workers, especially female workers (e.g. Jourdain, Chapter 7; Gajewski, Chapter 8). Finally, the book also considers how those workers at the margins 10 of the new economy—that is, those lacking the bodies, skills, and attitudes that define someone as employable—make a living and feel about it (Part IV). The first of the book’s five sections, titled ‘Trends and Scenarios in the Future of Work’, sets the general framework. It considers technological change as a historical trend within, or in relation to, other macro-social developments affecting work and social life more broadly. Part II, on ‘Digital Transforma- tions’, takes a closer look at the effects of digitalization on different areas of the economy and its implications for work. Part III, titled ‘Platform Work’, examines digital platforms and their impacts upon the timing, experience, and distribution of work. It shows how digitalization increases workloads in surreptitious ways. Part IV looks at the perceptions, choices, and experienc- es of marginalized workers, from elderly workers’ decision to retire or keep working based on their income levels to the feeling of shame jobseekers with disabilities endure when confronted with the welfare bureaucracy. Last but not least, Part V of the book examines the labour market trajectories and the future of work for researchers in and beyond academia. The remainder of this introduction synthesizes the contributions of each section to its subfield. By offering a socially embedded picture of the ongoing transformation of work and employment relations along multiple dimensions, from technology to welfare, and by discussing ways in which these dimensions intersect with each other, The Future of Work makes a timely contribution to understand the role of humans in 21st century capitalism. 1.2 Trends and Scenarios of the Future of Work Conventional wisdom associates digitalization with automation and the disappearance of paid employment opportunities. Some research supports this view. According to Frey and Osborne (2017), 47 percent of jobs in the US are at risk of automation. A report by McKinsey estimates that ‘between 400 million and 800 million individuals could be displaced by automation and need to find new jobs by 2030’ (Manyika et al. 2017). To the concern that machines will replace human bodies in the execution of routine phys- ical tasks—a concern as old as the industrial revolution itself—digitaliza- tion adds the risk of substitution in cognitive tasks as well as, increasingly, decision-making (Harari 2018). The ‘smart machine’—as Zuboff (1988) described computers in her classic study—‘informates’ production, that is, generates data about the process itself that can be used to further rationalize work. The skills originally embodied in workers are thus isolated, analysed, and, over time, replicated and transferred to non-humans. In this context, 11 Collins (2013) suggests, only the top 20 percent of students are likely to find a job. If so, for most of us, the future of work is joblessness. Another body of work emphasizes the degradation of employment across sectors and geographical areas—but tends to leave digital technology in the background. Precarity is the watchword of this literature. In Precarious Work, Kalleberg and Vallas (2017) present the rise of non-standard work arrangements such as temporary work, part-time work, on-demand work, and independent contracting as a secular trend spreading across the Global North. These arrangements replace the ‘good jobs’ (Kalleberg 2011) of the previous era, characterized by a living wage, full-time employment, fringe benefits (e.g. age-old pensions, health insurance, paid sick leave, and paid annual leave), and promotion perspectives (see also Kalleberg 2000). As the edited volume by Armano et al. (2017) makes plain, the wave of precariza- tion touches as much Japan as Greece or postcolonial Africa, and it affects precarious workers’ material conditions of existence as much as their subjec- tivities. In The Precariat, Guy Standing (2011) considers the political identity of this ‘new class’ that enjoys neither security of employment nor security of income. Some studies draw a link between digitalization and precarization, but, like Gray and Suri’s (2019) research on ‘ghost work’, usually focus on a specific set of occupations, such as Uber drivers or workers on TaskRabbit (Ravenelle 2019). Part I of this book addresses the current historical moment in the re- structuring of work and labour markets by connecting digitalization to other macro-social trends. Chapter 2 by Dominique Méda (in French) highlights the policy dimensions of precarization which, alongside technological devel- opment, contribute to emptying work of its meaningfulness and turn sensing human beings into abstract labour. Méda takes a long historical view to trace the evolution of the concept of work and discusses, against this background, the current challenges stemming from the brewing ecological crisis and, more recently, the Covid-19 pandemic. She outlines an alternative vision of work in society, which she terms reconversion écologique, and cites efforts to use the window of opportunity opened by the pandemic to enact lasting change. In Chapter 3, Chris Tilly considers, for his part, the dual processes of digitalization and informalization—which, like precarization, refers to the spread of unsafe, insecure, and uncertain forms of employment. Tilly’s chapter complements Méda’s historical view with a conceptual inquiry into the forces driving the disenfranchisement of workers. He puts forward a theoretical framework that stresses the embeddedness of employment in institutional structures as well as the role of shifting power relations and learning processes, all of which contribute to job degradation not only by themselves but through 12 their imbrication with technological change. In his final remarks, Tilly also points to avenues for collective action and resistance against these trends. 1.3 Digital Transformations Digitalization is spontaneously associated with manufacturing as technology affects, first and foremost, material processes of production. However, even the most casual survey of both the literature and real-life practices finds that digital technologies profoundly reshape organizational structures as well. For example, the corporate practice of outsourcing—both domestically through the use of contractors and internationally through global value chains (Ge- reffi 2018)—is facilitated by a dramatic fall of coordination costs thanks to new information and communication technologies. The old figure of the manager-foreman whose job was to ensure that employees did their work by—physically—supervising them has become obsolete in a computerized workplace where all activities performed by workers are or can be recorded, scrutinized, rated, and where workers themselves can easily be tracked and monitored in real-time, regardless of their location (Borja 2015). On the other hand, the effects of new technologies are not uniform across fields or industries. The same tool—e.g. touchscreens, barcodes, spreadsheets, or credit cards—can have widely different impacts depending on where they are introduced, for what purposes, and who uses them (Guseva and Rona-Tas 2014). Of particular import to understand the different impacts of innovation is the role of technology in reshaping the interface between the firm and its customers, be it consumers or other businesses. Indeed, the technology-driven restructuring of economic organizations happens not only through a mandated revision of the organigram by higher management seeking to cut costs but also organically, through an incremental or qualitative change in how products are made and delivered to customers. And whether products are purchased (and delivered) online, as in the case of Amazon, in brick-and-mortar stores, or both, makes a significant difference. Hence, Part II of this volume seeks to ground digitalization in more circumscribed areas of economic life. Chapter 4, by George Ritzer and Piergiorgio Degli Esposti, looks at relations with customers. It considers the partial realization of the end of (paid) work through the original lens of ‘prosumption’, a hybrid term combining ‘production’ and ‘consumption’. The authors argue that, thanks in part to digital innovation, work is not so much disappearing as it is being surreptitiously passed on to consumers, who are now increasingly—and oftentimes unwittingly—required to carry out 13 tasks formerly assigned to employees, from check-ins at airports to scanning groceries at the supermarket. Chapter 5, by Ulrich Dolata, focuses on platform economies. Dolata nuances the idea that digital technologies—and platform economies in par- ticular—revolutionize the organization of economic life; at the same time, he points to their broader impact on the structure of markets and on the rules and content of social exchanges, which platform operators define through their policies and algorithms. 1.4 Platform Work To fully grasp the implications of digital change for the world of work, Part III focuses on the distinct forms of employment that have emerged in the wake—or in the shadows—of the platform economies. A recurring question in the academic debate on the digitalization of work concerns the costs or benefits of ‘flexibility’. Echoing the broader polemic about the opportunities and hazards brought about by deregulated labour markets, digital platforms raise various questions about control over, and the freedom of, workers in the gig economy (Ravenelle 2019; Spreitzer et al. 2017). In The Future of Crowd Work, Kittur et al. (2012) point to the opportunities that this internet-based organization of work offers ‘for improving produc- tivity, social mobility, and the global economy by engaging a geographically distributed workforce to complete complex tasks on demand and at scale.’ It is also the case that some workers with skills in high demand such as com- puter programming can take advantage of their independence as freelancers to decide when to work and for whom or to book last-minute vacations to Hawaii (Kessler 2018). A paper by Chen et al. (2019, 1) argues that ‘Uber drivers benefit significantly from real-time flexibility, earning more than twice the surplus they would in less flexible arrangements.’ In fact, the most enthusiastic supporters paint digital platforms as ‘entrepreneurial incubators’, as Vallas and Schor (2020) note in a recent review. By contrast, critical perspectives present gig workers as exploited mem- bers of a postindustrial proletariat that has lost the stability of employment it enjoyed under industrial capitalism—a trend captured by neologisms like ‘flexploitation’ (see Bourdieu 1998) or ‘precariat’ (see Standing 2011). In fact, the EU parliamentary report by Forde et al. (2017) finds that workers in digitally mediated employment earn in most cases less than the minimum wage. And more gig workers are likely to hold multiple jobs than estimated by standard workforce surveys (Katz and Krueger 2019). Given high rates 14 of market concentration in platform economies, moreover, gig workers who derive the bulk of their income from a single platform belong to the expanding category of dependent self-employed workers or faux indépendants (Bernard 2020; Bonvin 2017). This relationship of implicit dependence can have serious consequences because, even if they do not present themselves as employers, platform owners retain ‘the right to supervise, discipline or discharge the worker or prevent their use of the platform’ (Stewart and Stanford 2017, 5). Indeed, platform workers cannot be fired, but they can be ‘deactivated’ (Kessler 2018). Hence the troubling description of the digitalized world of work as an algorithmic tyranny (Harari 2018). Contributions to Part III of this volume paint a more complex and nuanced picture by showing that platform workers, while sometimes adhering to the narrative of flexibility, end up doing more work, paid and unpaid, for the benefit of others, and that digital technologies allow companies to transfer a share of work to consumers. Chapter 6 by Anne Jourdain studies the effects of the handcraft e-commerce platform Etsy on women’s work. Jourdain finds that instead of realizing the promise of making money while pursuing their hobbies, female workers on Etsy end up performing more unpaid, invisible work both for the platform and in their households. Indeed, women who enjoy the ‘privilege’ of working from home are often compelled to take care of household chores during the day and still depend on their husbands for income. Chapter 7 by Eltje Gajewski studies crowd workers’ own understandings of their work and employment status. Gajewski finds that crowd workers adhere to the narrative of flexibility even as they experience tight constraints on their schedules and work routines. Workers dedicate more and more time to work and obsessively focus on improving their performances, which suggests platforms achieve as much if not more worker discipline at a much lower cost than the traditional, hierarchical, bureaucratic firm. 1.5 Work at the Margins The globalization of labour markets coupled with skill-biased technical change and automation breed a surplus workforce that struggles to find employment in both the Global South and North. Yet the exclusion of workers is not only enacted through structural processes. Indeed, the world of work is organized materially—through the production techniques that digital technologies and globalization have disrupted—but also symbolically and institutionally, by the system of social categories that employers and the state apply to workers 15 and their work. A well-established tradition in the sociology of professions studies mechanism of social closure that bar potential competitors from certain types of work (Weeden 2002). Among the groups discriminated against or marginalized in the new economy are the youth (Weil et al. 2005), the elderly (Bouffartigue et al. 2017), foreigners, people of colour (Pager et al. 2009), and jobseekers with few years of schooling, no professional training, and/or lacking educational credentials (Blossfeld 1983; Bynner and Parsons 2002). Women feature prominently among groups historically excluded from the labour market and unable to access the material and symbolic benefits associated with a career. Exclusion from paid employment, either by law or social convention, long co-existed with the undervaluation of house chores performed by women, subsequently conceptualized as ‘invisible work’ (Ar- borio 2017; Kergoat 2001). Starting in the 1960s, changes in legislation, public policy, and social attitudes enabled women to join the labour force in large numbers and to gain access to a widening range of occupations (Simms 2019). Nevertheless, closure leaves a long-term imprint, and the social divi- sion of labour is still shaped by widespread gender stereotypes and gendered practices around work (Buchmann et al. 2003; Ridgeway 2011; Le Feuvre et al. 2015; Ferrary 2018). The historical experience of women points to the role of the state as an agent of integration into the labour market. Its protective role involves not only the adoption and enforcement of labour protection and gender equal- ity laws, however, but also the provision of sustenance to those temporarily or permanently excluded from the labour market. On the other hand, the decommodification of labour is jeopardized by welfare state retrenchment, as eligibility requirements for social assistance become tighter and are tied to compulsory job searches (a policy known as ‘workfare’). In this context, it is usually assumed that socially stigmatized, disadvantaged job-seekers resort to precarious self-employment or ‘end up in low-status, low-paid, and insecure jobs’ (Gesthuizen et al. 2011, 264; see also Schilling et al. 2019). Working off the books is a common survival strategy—including among welfare recipients (Edin and Lein 1997)—identified in the literature (Venkatesh 2006). And economic informality is associated with lack of labour rights and subpar labour standards in developing countries—a problem aggravated by globalization (Centeno and Portes 2006, Cuvi 2019). Part IV considers alternative coping strategies along with the categoriza- tion and feelings of three disadvantaged groups: ageing low-income workers, low-skill workers, and people with disabilities. It uncovers the material prac- tices and the subjective experience of marginalization. In Chapter 8, Rainer Gabriel and colleagues present a longitudinal study of poverty in old age in 16 Switzerland and examine causal factors behind both income-poverty and subjective poverty among people over sixty-five. They find that employment status in the first job is a strong predictor of both. Workers whose first job was blue-collar are more likely to fall into poverty during their old age. Finally, while early retirement is prevalent among workers with incomes above the poverty line, so is working after retirement. By using panel data from the largest gerontology survey in Switzerland, Gabriel and colleagues are thus able to isolate novel patterns in the relation between work and poverty. Chapter 9 by Eva Nadai and Anna Gonon looks at the discourse and practices of management surrounding ‘low-skill’ jobs and workers in firms. They find that, notwithstanding a general discourse about the disappearance of low-skill jobs, employers in various sectors still need and hire workers with little or no formal qualifications for jobs employers cast as ‘simple’. Instead of recognizing those skills that they look for as qualifications for the job, managers frame them as personal qualities that some workers naturally have and that enable them to do work ‘that everybody can do’. By analysing these processes through the economy of conventions theory, the authors thus provide original insights into the logic of worker and skill valuation, which is far more complex than a cold assessment of individual worker productivity. Jean-Pierre Tabin and his co-authors study in Chapter 10 (in French) the experience of workers in Switzerland identified as disabled—that is, according to Swiss law, ‘persons whose earning capacity is diminished because of a health impairment’ (Tabin et al. 2019, 1). These people describe their experiences through the language of shame associated with a status of inferiority embed- ded in the ableist vision endorsed by welfare institutions and society at large. The chapter thus teases out what it means to not work, or to not be able to work, and the multiple ways—from being suspected of taking advantage of the system to being sent back to school—in which those in that condition are told that they deviate from the norm. 1.6 The Future of Work in Academia The dual processes of digitalization and precarization affect the work of scholars in many ways. For one, they change the methods of research quite substantially, as data becomes available in (digital) formats even quantitative sociologists are not always familiar with. Big data, in particular, offer countless opportunities to map and probe society, yet they require skills in recently cre- ated and constantly evolving languages, such as R or Python. Boasting about the social behaviours his colleagues were able to predict based on analysis of 17 Google searches, a computer scientist recently told one of the authors of this introduction that sociologists found themselves in ‘an arms race [they] cannot win’ and suggested collaborations as a possible way forward. On the other hand, social scientists have the potential to turn a theoretically informed, critical eye to the use of these tools and the results derived from them, as Noortje Marres (2017) argues in Digital Sociology. They can also illuminate through traditional methods how these tools are applied by, and change the work of, other actors and organizations (Brayne 2017). At a deeper level, Mike Savage (2015) recently made the case that social change related to or driven by new technologies is also changing what the social sciences look like and the kind of work that gets attention from the public. Citing the success of Piketty’s (2014) Capital in the Twenty-First Cen- tury and Wilkinson and Pickett’s The Spirit Level (2009), Savage noted that highly sophisticated theoretical endeavours of the kind that made famous the great social thinkers of the 20th century have lost currency in favour of works that tease out new trends in rich quantitative databases.1 The jury is still out as to the devaluation of theoretical thinking in the social sciences and the role, if any, that digital technologies may be playing in it. But the need for creative and ambitious interpretative frameworks to make sense of the world in these unsettled times is undeniable. More immediately worrisome, at least to the next generation of academics, is the lack of secure employment prospects. Not only is such insecurity of employment likely to deter promising candidates from pursuing an academic career, but the instability to which those who do so are subject to will affect the quality and volume of their research. The freedom to think detached from the urgency of procuring a livelihood is key to the making of scholars. As Standing (2011) argues, moreover, the material conditions of existence that characterize the precariat have long-term detrimental cognitive effects. If precarization is indeed a structural trend, as Tilly (this volume) and others suggest, then the scholarly mind may be a thing of the past. Fortunately, academics are taking a stance against precarity across Europe and beyond, which suggests the issue is a global one (Neveu and Surdez 2020). However, its effects play out differently based on national and, in some cases, subnational contexts. In Switzerland, where the congress that gave birth to this volume took place, the massification of higher education took place against the background of a vertically structured system of professorial chairs, which remained in place. As a result, professorships are few, and many of the duties traditionally associated with them—including research and teaching—are 1 Mike Savage would subsequently nuance this claim, which did not go down well with his German audience. 18 performed by a large body of mid-level academic workers whose prospects of securing a permanent job in their home country are remote, to say the least. Part V of this book examines the job market for Swiss academics at a time of both precarization, limited career opportunities in academia, and growing international exchanges. Chapter 11, by Pierre Bataille and Marie Sautier (in French), analyses the trajectories of PhD-holders in Switzerland who seek to pursue an academic career. Faced with the internationalization of the academic job market and research networks as well as the expectation of international experience from universities and funding agencies in Switzerland, these young researchers end up leaving the country where they would like to live and work, and they are sometimes unable to come back. Based on results from a survey conducted in Switzerland, Chapter 12 by Rainer Diaz-Bone looks at what happens for those who stay in Switzerland but leave the academic field. Diaz-Bone argues that sociologists have been relatively successful at obtaining non-academic jobs in both the private and public sectors despite not having a clear career path associated to their back- ground outside academia. 1.7 Back to the Future? As the world came to terms with the scale of the Covid-19 crisis in early 2020 and governments scrambled to contain the spread of the virus by imposing drastic social distancing measures such as curfews and lockdowns, a line had to be drawn between workers deemed ‘essential’, who were allowed and even compelled to go to work, and others, who were instructed to stay home. As a result of the Covid-19 lockdown measures internet-based platform work has certainly increased (cf. OFS 2020b for data on the situation of platform work in Switzerland before the crisis). For many, mostly white-collar workers, technology came to the rescue. Computer devices and collaborative software allowed them to carry on with (part of ) their work routines and hold business meetings on Zoom, Webex, and the like. Telecommuting during the pandemic came at a price, however, as working parents had to grapple with the challenges of home-schooling and a lack of childcare. Still, the stress endured by these workers pales in comparison to the woes of millions of other workers who simply lost their jobs. In an unprecedented development, more than 6 million people filed for unemployment benefits in the United States in the third week of March, and then another 6.6 million the week after, bringing the unemployment rate to a historic high (Labor Department 2020). In a country where health 19 insurance is often provided by the employer, a job loss compromised access to healthcare for many workers at the worst possible time while also jeopardizing access to other basic services, such as housing. ‘No job, no rent’ read placards hanging from the windows of desperate tenants in the Bronx. Essential workers kept their jobs (and mobility) but faced another set of challenges. As cases spiked at meatpacking plants and warehouses, some em- ployees staged protests demanding protective equipment from their employer. Not all essential workers had health coverage, moreover. The steady increase in non-standard work arrangements and management strategies aimed at cutting labour costs had led to an erosion of employment-related benefits, including health insurance (Kalleberg and Vallas 2017). In the Global South, dependence on informal economic activity carried out on public spaces, such as street trade, or in crowded indoor spaces confronted a majority of the workforce to the tragic dilemma of having to either work and risk catching Covid-19 or go hungry. In April 2020, the International Labour Organization estimated that 1.6 billion workers in the informal economy were at risk of losing their livelihoods (ILO 2020). In many ways, therefore, the Covid-19 pandemic laid bare deep-running problems with the structure of labour markets and the lack of social pro- tection. As contributions to this book make plain, the re-commodification of labour involves not only the disenfranchisement of workers but also a breakdown of work into smaller tasks paid at piece rate and the extension and flexibilization of work schedules enabled by digital platforms (see also Cianferoni 2019). It erases boundaries that used to protect workers. At the limits, the re-commodification of labour has the paradoxical effects of forcing a share of the workforce out of the economy, as Saskia Sassen (2014) suggests in her book Expulsions, and of fostering unpaid work in institutions such as prisons or rehab centres. In both the post-industrial North and the deindustrializing South, however, the Covid-19 crisis sparked calls for the reinstatement of social protections for frontline workers, laid-off workers, and independent contractors unable to work (Suzman 2020). From Brazil to Malaysia, governments enacted social programs aimed at providing economic relief to formal and informal workers. Under the CARES Act, gig workers became eligible for unemployment insur- ance in the United States. The crisis also triggered a collective recognition of value in the work of others, the most dramatic expression of which were the daily rounds of applauses and pan-banging to acknowledge frontline workers. In fact, signs of a double-movement—to use Polanyi’s ([1944] 2001) famous 20 phrase referring to backlash against the expanding influence of market forces in society—preceded the Covid-19 crisis.2 And academic workers felt the pinch. Educators forced to teach online soon discovered the practical and pedagogical challenges of interacting with students on screens, often from each other’s houses. Qualitative researchers had to rethink their fieldwork strategies in a context of mandatory social distancing and weigh the pros and cons of conducting Skype interviews. International conferences were cancelled, postponed, or moved online. These challenges added to the market-related stresses already facing the new (and not so new) generations of scholars, which have seen job opportunities and the terms of employment deteriorate in recent years (ASSH 2018). In the United States, budget cuts at universities following the lockdown led many graduating and soon-to-graduate researchers to consider employment outside of academia, and several top sociology departments froze admissions into their graduate programs. May the current crisis inspire a rethink of traditional structures and foster a more equitable organization of the field. 1.8 References Arborio, Anne-Marie. 2017. ‘Visibilisation, invisibilisation du travail. Introduc- tion.’ In José-Angel Calderon, Lise Demailly and Séverin Muller (Eds.), Aux marges du travail. Toulouse : Octares, 11–13. Armano, Emiliana, Arianna Bove, and Annalisa Murgia. 2017. Mapping Precari- ousness, Labour Insecurity and Uncertain Livelihoods: Subjectivities and Resistance. London: Taylor & Francis. ASSH. 2018. Next Generation: pour une promotion efficace de la relève. Swiss Academies Reports, 13(2): 1–71. https://doi.org/10.5281/ZENODO.3662439. Baldwin, Richard. 2006. Globalisation: The Great Unbundling(s). Prime Minister’s Office: Economic Council of Finland. 20 September, 2006, http://appli8. hec.fr/map/files/globalisationthegreatunbundling(s).pdf (accessed February 15, 2021). Barbarrusa, Tomás Gutiérrez. 2016. La montée du travail précaire en Europe. Concepts, indicateurs et effets de la crise économique mondiale. Revue Inter- nationale du Travail, 155(4): 525–559. https://doi.org/10.1111/ilrf.12033. 2 In late 2019 and early 2020, France was paralysed by strikes against pension reform. The business community itself appeared to take notice of the dangers associated with unbridled labour markets when the CEOs of the largest corporations in the US, grouped under the umbrella of the Business Roundtable, pledged to embrace a more comprehensive approach to management that considers the well-being of workers and consumers instead of focusing exclusively on maximizing shareholder value. 21 Bernard, Sophie. 2020. Le nouvel esprit du salariat : Rémunérations, autonomie, inégalités. Paris: Presses universitaires de France. Bettio, Francesca, Jenneke Plantenga, and Mark Smith (Eds.). 2013. Gender and the European Labour Market. London and New York: Routledge. Bonvin, Jean-Michel. 2017. Nous ne sommes pas sortis de la société du salariat. Panorama : Formation, Orientation, Marché du Travail, 4: 6‒7. Borja, Eric Enrique. 2015. ‘Chip: The Costs of Chasing the American Dream.’ In J. Auyero (Ed.), Invisible in Austin: Life and Labor in an American City. Austin: University of Texas Press, 99‒115. Bouffartigue, Paul, Mariana Busso, and Marcos Supervielle. 2017. ‘Introduction.’ In Paul Bouffartigue, Mariana Busso, and Marcos Supervielle (Eds.), Travail, jeunesse et migrations : Regards croisés Europe-Amérique latine à l’heure de la mondialisation. Paris : Éditions de l’IHEAL, 7–22. Bourdieu, Pierre. 1998. La précarité est aujourd’hui partout. Contre-feux. Paris : Raison d’Agir, 95‒101. Brayne, Sarah. 2017. Big Data Surveillance: The Case of Policing. American Sociological Review, 82(5), 977‒1008. Buchmann, Marlis, Irene Kriesi, and Stefan Sacchi. 2003. Labour-Market Struc- tures and Women’s Paid Work: Opportunities and Constraints in the Swiss Labour Market. Advances in Life Course Research, 8: 165–188. https://doi. org/10.1016/S1040-2608(03)08008-0. Bühlmann, Félix. 2013. Catch Me if You Can. Configurations of Atypical and Precarious Employment in Switzerland. Espace populations sociétés. Space populations societies, 3: 73–90. https://doi.org/10.4000/eps.5554. Bynner, John and Samantha Parsons. 2002. Social Exclusion and the Transition from School to Work: The Case of Young People Not in Education, Employ- ment, or Training (NEET). Journal of Vocational Behavior, 60(2): 289–309. https://doi.org/10.1006/jvbe.2001.1868. Cappelli, Peter, Laurie Bassi, Harry Katz, David Knoke, Paul Osterman, and Michael Useem. 1997. Change at Work. New York and Oxford: Oxford Uni- versity Press. Centeno, Miguel Angel and Alejandro Portes. 2006. ‘The Informal Economy in the Shadow of the State.’ In Patricia Fernández-Kelly and Jon Shefner (Eds.), Out of the Shadows: Political Action and the Informal Economy in Latin America. Pennsylvania State University Press. Cianferoni, Nicola. 2019. Travailler dans la grande distribution : La journée de travail va-t-elle redevenir une question sociale ? Zurich et Genève : Éditions Seismo. Chen, M. Keith, Peter E. Rossi, Judith A. Chevalier, and Emily Oehlsen. 2019. The Value of Flexible Work: Evidence from Uber Drivers. Journal of Political Economy, 127(6): 2735‒2794. 22 Collins, Randall. 2013. ‘The End of Middle Class Work: No More Escapes.’ In Immanuel Wallerstein, Randall Collins, Michael Mann, Georgi Derluguian, and Craig Calhoun (Eds.), Does Capitalism Have a Future? Oxford: Oxford University Press, 37–70. Cuvi, Jacinto. 2019. Symbolic Capital, Informal Labor, and Postindustrial Markets: The Dynamics of Street Vending During the 2014 World Cup in São Paulo. Theory and Society, 48(2), 217–238. Edin, Kathryn and Laura Lein. 1997. Making Ends Meet: How Single Mothers Survive Welfare and Low-Wage Work. New York: Russell Sage Foundation. Ehrler, Franziska, Felix Bühlmann, Peter Farago, François Höpflinger, Dominique Joye, Pasqualina Perrig-Chiello, and Christian Suter (Eds.). 2016. Swiss Social Report 2016: Wellbeing. Zurich: Seismo. Ferrary, Michel. 2018. Gender Diversity in the Labour Market: Employer Discrimination, Educational Choices and Professional Preferences. @GRH, 27(2): 83–118. Forde, Chris, Mark Stuart, Simon Joyce, Liz Oliver, Danat Valizade, Gabriella Alberti, and Calum Carson. 2017. The Social Protection of Workers in the Platform Economy. Brussels: Policy Department-Economic and Social Policy, Employment and Social Affairs (European Parliament). November, 2017, https://www.europarl.europa.eu/RegData/etudes/STUD/2017/614184/ IPOL_STU(2017)614184_EN.pdf (accessed February 15, 2021). Frey, Carl Benedikt and Michael A. Osborne. 2017. The Future of Employment: How Susceptible Are Jobs to Computerisation? Technological Forecasting and Social Change, 114: 254–280. https://doi.org/10.1016/j.techfore.2016.08.019. Gatta, Mary. 2014. All I Want is a Job!: Unemployed Women Navigating the Public Workforce System. Stanford, CA: Stanford University Press. Gereffi, Gary. 2018. Global Value Chains and Development: Redefining the Contours of 21st Century Capitalism (Development Trajectories in Global Value Chains). Cambridge: Cambridge University Press. Gesthuizen, M., H. Solga, and R. Kunster. 2011. Context Matters: Economic Marginalization of Low-Educated Workers in Cross-National Perspective. European Sociological Review, 27(2): 264–280. https://doi.org/10.1093/esr/ jcq006. Gray, Mary L. and Siddharth Suri. 2019. Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass. Boston: Houghton Mifflin Harcourt. Guseva, Alya and Akos Rona-Tas. 2014. Plastic Money: Constructing Markets for Credit Cards in Eight Postcommunist Countries. Standford, CA: Stanford University Press. Harari, Yuval Noah. 2018. Why Technology Favors Tyranny. The Atlantic, 322(3), October, 2018, https://www.theatlantic.com/magazine/archive/2018/10/ yuval-noah-harari-technology-tyranny/568330/ (accessed February 15, 2021). 23 International Labour Organization. 2020. ‘ILO Monitor: COVID-19 and the world of work. Third edition. Updated estimates and analysis’ published April 29. https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/ documents/briefingnote/wcms_743146.pdf (accessed July 30, 2020). Kalleberg, Arne L. 2000. Non-Standard Employment Relations: Part-time, Temporary and Contract Work. Annual Review of Sociology, 26(1): 341‒365. Kalleberg, Arne L. 2011. Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s‒2000s. New York: Russell Sage Foundation. Kalleberg, Arne L. and Steven P. Vallas. 2017. Precarious Work. Bingley: Emerald Group Publishing. Kergoat, Danièle. 2001. ‘Division Sexuelle Du Travail et Rapports Sociaux de Sexe.’ In Jeanne Bisilliat and Christine Verschuur (Eds.). Genre et économie : un premier éclairage. Genève : Graduate Institute Publications, 78–88. Katz, Lawrence F., and Alan B. Krueger. 2019. ‘The Rise and Nature of Alternative Work Arrangements in the United States, 1995–2015.’ ILR Review, 72(2) (March): 382–416. https://doi.org/10.1177/0019793918820008. Kessler, Sarah. 2018. Gigged: The End of the Job and the Future of Work. New York: St. Martin’s Publishing Group. Kittur, Aniket, Jeffrey V. Nickerson, Michael Bernstein, Elizabeth Gerber, Aaron Shaw, John Zimmerman, Matt Lease, and John Horton. 2012. The Future of Crowd Work. Proceedings of the 2013 Conference on Computer Supported Coopera- tive Work, CSCW ’13, 1301–1318. https://doi.org/10.1145/2441776.2441923. Labour Department. 2020. ‘Seasonally Adjusted US Weekly UI Claims’ News Release, July 23. https://www.dol.gov/ui/data.pdf (accessed July 30, 2020). Lasi, Heiner, Peter Fettke, Hans-Georg Kemper, Thomas Feld, and Michael Hoffmann. 2014. Industry 4.0. Business & Information Systems Engineering, 6(4): 239–242. https://doi.org/10.1007/s12599-014-0334-4. Le Feuvre, Nicky, Morgane Kuehni, Magdalena Rosende, and Céline Schoeni. 2015. ‘Gendered Variations in the Experience of Ageing at Work in Swit- zerland.’ In Brigitte Liebig and René Levy (Eds.), Equality, Diversity and Inclusion: An International Journal, 34(2): 168–181. https://doi.org/10.1108/ EDI-03-2014-0017. Manyika, James, Susan Lund, Michael Chui, Jacques Bughin, Jonathan Woetzel, Parul Batra, Ryan Ko, and Saurabh Sanghvi. 2017. Jobs Lost, Jobs Gained: Workforce Transitions in a Time of Automation. McKinsey Global Institute. December, 2017, https://www.mckinsey.com/~/media/McKinsey/Industries/ Public%20and%20Social%20Sector/Our%20Insights/What%20the%20 future%20of%20work%20will%20mean%20for%20jobs%20skills%20 and%20wages/MGI-Jobs-Lost-Jobs-Gained-Executive-summary-Decem ber-6-2017.pdf (accessed February 15, 2021). 24 Marres, Noortje. 2017.  Digital Sociology: The Reinvention of Social Research. London: John Wiley & Sons. Neveu, Erik and Muriel Surdez. 2020. Globalizing Issues. London: Palgrave McMillan. OCDE. 2014. Perspectives de l’emploi de l’OCDE. Éditions OCDE, https://doi. org/10.1787/empl_outlook-2014-fr (accessed February 15, 2021). OFS. 2006. Le travail à temps partiel en Suisse. Neuchâtel : Office fédéral de la statistique. 4 Août, 2006, https://www.bfs.admin.ch/asset/fr/206-0608 (accessed February 15, 2021). OFS. 2020a. Indicateurs du marché du travail 2020. Neuchâtel : Office fédéral de la statistique. 11 Août, 2020, https://www.bfs.admin.ch/bfs/fr/home/ statistiques/travail-remuneration/travail-non-remunere/compte-satellite-pro- duction-menages.assetdetail.13627183.html (accessed February 15, 2021). OFS. 2020b. Le travial via les plateformes numériques est peu répandu en Suisse. Communiqueé de presse. Neuchâtel : Office fédéral de la statistique OFS. 19 May, 2020, https://www.bfs.admin.ch/bfs/fr/home/statistiques/ travail-remuneration/activite-professionnelle-temps-travail/actifs-occupes/ conditions-travail/services-plateformes-numeriques.assetdetail.12787863. html (accessed March 21, 2021). Pager, Devah, Bart Bonikowski, and Bruce Western. 2009. Discrimination in a Low-Wage Labour Market: A Field Experiment. American Sociological Review, 74(5): 777–799. https://doi.org/10.1177/000312240907400505. Piketty, Thomas. 2014. Capital in the Twenty-First Century. Cambridge and Lon- don: Harvard University Press Polanyi, Karl. 2001 [1944]. The Great Transformation: The Political and Economic Origins of Our Time. Boston: Beacon Press. Preoteasa, Ana Maria, Rebekka Sieber, Monica Budowksi, and Christian Suter. 2016. Household Role in Coping with Precarious Work. Evidence from Qualitative Research in Urban Romania and Switzerland. Social Change Review, 2: 177‒201. Raess, Damian and Brian Burgoon. 2015. Flexible Work and Immigration in Europe. British Journal of Industrial Relations, 53(1): 94‒111. Ravenelle, Alexandrea J. 2019. ‘We’re not uber:’ Control, Autonomy, and En- trepreneurship in the Gig Economy. Journal of Managerial Psychology, 34(4): 269‒285. Ridgeway, Cecilia L. 2011. Framed by Gender: How Gender Inequality Persists in the Modern World. New York: Oxford University Press. Sassen, Saskia. 2014. Expulsions. Brutality and Complexity in the Global Economy. Cambridge and London: Harvard University Press. 25 Savage, Mike. 2015. ‘Making Sense of Big Data’ Presentation at the British Em- bassy in Berlin. https://www.youtube.com/watch?v=9urYbd_rVoc (accessed February 15, 2021). Schilling, Hannah, Talja Blokland, and AbdouMaliq Simone. 2019. Working Precarity: Urban Youth Tactics to Make Livelihoods in Instable Conditions in Abidjan, Athens, Berlin and Jakarta. The Sociological Review, 67(6): 1333–49. https://doi.org/10.1177/0038026119858209. Silver, Beverly J. 2003. Forces of Labour: Workers’ Movements and Globalisation since 1870. Cambridge: Cambridge University Press. Simms, Melanie. 2019. The Future of Work. What Do We Know and What Should We Do About. Los Angeles: SAGE. Stiglitz, Joseph E. 2001. Globalisation and Its Discontents. New York: W.W. Norton & Company. Spreitzer, Gretchen M., Lindsey Cameron, and Lyndon Garrett. 2017. Alternative Work Arrangements: Two Images of the New World of Work. Annual Review of Organizational Psychology and Organizational Behavior, 4(1): 473–499. https://doi.org/10.1146/annurev-orgpsych-032516-113332. Standing, Guy. 2011. The Precariat: The New Dangerous Class. London: Blooms- bury Publishing. Stewart, Andrew and Jim Stanford. 2017. Regulating Work in the Gig Economy: What Are the Options? The Economic and Labour Relations Review, 28(3): 420–437. https://doi.org/10.1177/1035304617722461. Susskind, Richard E. and Daniel Susskind. 2015. The Future of the Professions: How Technology Will Transform the Work of Human Experts. Oxford: Oxford University Press. Suter, Christian. 2009. Inequality Beyond Globalisation: Searching for the Miss- ing Pieces of the Puzzle: Introduction to Special Issue. International Journal of Comparative Sociology, 50(5‒6): 419‒424. Suzman, James. 2020. The 300,000-year Case for the 15-Hour Week. Financial Times, 28 August, 2020, https://www.ft.com/content/8dd71dc3-4566-48e0- a1d9-3e8bd2b3f60f (accessed February 15, 2021). Tabin, Jean-Pierre, Monika Piecek, Céline Perrin, and Isabelle Probst. 2019. Three Dimensions in the Register of Shame. Review of Disability Studies: An International Journal, 15(3): 1‒19. Vallas, Steven P. and Juliet B. Schor. 2020. What Do Platforms Do? Understanding the Gig Economy. Annual Review of Sociology, 46(1): 273‒294. Venkatesh, Sudhir Alladi. 2006. Off the Books. Cambridge and London: Harvard University Press. Weeden, Kim A. 2002. Why Do Some Occupations Pay More than Others? Social Closure and Earnings Inequality in the United States. American Journal of Sociology, 108(1): 55–101. https://doi.org/10.1086/344121. 26 Weil, Susan Warner, Danny Wildemeersch, and Theo Jansen. 2005. Unemployed Youth and Social Exclusion in Europe: Learning for Inclusion? Aldershot: Ashgate. Wilkinson, Richard and Kate Pickett. 2009. The Spirit Level: Why Equality Is Better For Everyone. London: Penguin UK. Zuboff, Shoshana. 1988. In the Age of the Smart Machine: The Future of Work and Power. New York: Basic Books. Part I Trends and Scenarios of the Future of Work 29 2 Sens et avenir du travail en Europe Dominique Méda 2.1 Introduction Les historiens et les anthropologues semblent à peu près tous d’accord pour dire aujourd’hui que l’économique, la production, le travail... tels que nous les entendons dans nos sociétés sont des notions et des domaines qui se sont constitués clairement à partir du XVIIIe siècle en Europe, avec la différenciation d’un marché capitaliste au sein du marché qui lui préexistait. (Freyssenet 1999, 2) Ainsi s’exprimait, en 1999, le sociologue français Michel Freyssenet en défendant une thèse que je partageais mais qui, contrairement à ce qu’affirme l’auteur, était et reste encore controversée. C’est en prenant néanmoins appui sur elle que je m’interroge dans ce texte sur la manière dont il est possible de satisfaire les immenses attentes qui sont aujourd’hui placées sur le travail, notamment par les Européens. Je rappelle en effet dans un premier moment la place de plus en plus importante que le travail occupe aujourd’hui dans nos vies, à la fois gagne-pain, moyen d’expression, support de lien social, de droits et de protections. Je propose ensuite de distinguer plusieurs scénarios ou configurations qui caractérisent déjà et pourraient caractériser le futur du travail en examinant dans quelle mesure chacun·e est capable ou non de faire droit aux attentes de sens et de réalisation de soi qui sont actuellement fortement présentes chez les Européens. En m’appuyant sur les études dont nous pouvons déjà disposer, j’examine enfin les bouleversements que la pandémie de Covid-19 a entrainés et les inflexions de tendance sur lesquelles elle pourrait déboucher. 30 2.2 L’invention du travail Je voudrais donc commencer cette réflexion sur l’avenir du travail en pre- nant pour quelques instants une vue moins prospective que rétrospective de manière à insister sur un point qui me paraît absolument majeur : le caractère historique du concept de travail. Certes, les humains ont toujours cherché à organiser leur milieu, inventé des outils et radicalement transformé leur environnement – au point que nous nous demandons aujourd’hui si celui-ci pourra encore longtemps nous supporter. Mais les finalités de ces activités et les représentations qui les accompagnaient ont longtemps été radicalement différentes des nôtres, ce qui est perceptible si nous nous interdisons de lire le passé à l’aide de nos propres catégories. L’historien Jean-Pierre Vernant écrivait de manière lumineuse en 1965 : « de même qu’on n’a pas le droit d’appliquer au monde grec les catégories économiques du capitalisme moderne, on ne peut projeter sur l’homme de la cité ancienne la fonction psychologique du travail telle qu’elle est aujourd’hui dessinée » (Vernant 1996, 295). Exactement au même moment, dans ses travaux sur l’Algérie, Pierre Bourdieu mettait en évidence les fonctions radicalement différentes assurées par le travail dans une société précapitaliste bouleversée par la modernité capitaliste ; à plusieurs reprises dans son cours d’Anthropologie économique, il défend l’idée d’une « invention du travail », un terme qui sera fortement mobilisé dans les réflexions anthropologiques et sociologiques en Allemagne et en France dans les années 1980 et 1990. En 1982, le XXIe Congrès allemand de sociologie, qui se tient à Bamberg et s’ouvre sur le thème général de la crise de la société du travail, notamment avec des conférences plénières de Ralf Dahrendorf (Wenn der Arbeitsgesellschaft die Arbeit ausgeht) et de Claus Offe (Arbeit als soziologische Schlüsselkategorie ?) donne une place de choix aux théories qui envisagent le basculement d’une société du travail à une société d’activité. J’ai moi-même pris part à ce débat en France avec mon ouvrage Le travail. Une valeur en voie de disparition ? publié en 1995 (Méda, 1998). Je voudrais d’un mot rappeler ma thèse : notre concept moderne de travail est composé de plusieurs couches de signification qui ont été ajoutées par les différentes périodes historiques et qui se sont en quelque sorte sédimentées. Ce n’est qu’au XVIIIe siècle que la notion de travail trouve son unité – notamment au travers de la théorisation de Smith – unité qui se paye néanmoins par le caractère abstrait du concept. C’est l’invention du travail abstrait, du travail quantité divisible à l’infini, dont l’essence est réductible au temps et qui se présente avant tout comme ce que les économistes appellent aujourd’hui un « facteur de production ». Le travail se caractérise par sa dimension ins- 31 trumentale, il est « ce qui crée de la richesse ». Sur cette première couche de signification, le XIXe siècle amènera une dimension radicalement nouvelle, le travail-liberté, le travail « essence de l’homme », conçu comme puissance transformatrice du monde et de soi-même. L’humain a pour vocation de faire le monde à son image, et, comme le suggère Hegel selon Pierre Chamley (1965), d’anéantir (vernichten) la nature (ou ce qui est donné) pour faire advenir l’Esprit. C’est alors l’activité même de travail qui est déterminante, son contenu puissamment transformateur, créateur, émancipateur, comme le met en évidence le terme de Bildung plus souvent employé par Hegel. Marx représente l’acmé de ce courant de pensée lorsqu’il se moque de Smith (parce que ce dernier n’a pas compris que le travail est une passion créatrice, posi- tive) et écrit dans les Manuscrits parisiens : « supposons que nous produisions comme des êtres humains […] Nos productions seraient autant de miroirs où nos être rayonneraient l’un vers l’autre » (Marx 1979, 33). La fin du XIXe siècle voit enfin l’avènement, dans les sociétés occidentales, d’une troisième dimension : le travail comme support des droits et des protections caractéris- tiques de la société salariale. Tout au long de ces deux siècles pendant lesquels s’est construit le concept moderne de travail, celui-ci est décrit comme une activité profondément génératrice de lien social. Le magnifique texte qui clôt le premier chapitre des Recherches sur la nature et les causes de la richesse des nations montre dans son entrelacs même combien la réalisation d’une simple chemise requiert le concours d’un très grand nombre de personnes, dont Adam Smith (1999) a rappelé à de multiples reprises le penchant originel à troquer et échanger. La dialectique du maître et de l’esclave présente dans la Phénoménologie de l’Esprit illustre magistralement le fait que le travail est certes un acte individuel de mise en forme du donné mais aussi un rapport social. Emile Durkheim (2013) propose quant à lui de considérer la division sociale du travail au centre du fonctionnement social, en soulignant la fonc- tion éminemment morale du travail. Ces trois dimensions – le travail abstrait créateur de richesse, essence de l’homme et pivot de la société salariale – co-existent aujourd’hui, à des degrés divers, dans nos représentations du travail et ces variations sont déterminées par de nombreux éléments parmi lesquels notamment le niveau de revenu par habitant, de chômage et d’éducation. Et plus qu’une substitution des dimensions post-matérialistes aux dimensions instrumentales suggérée par les travaux d’Inglehart (1990), on constate aujourd’hui une coprésence des différentes aspirations vis à vis du travail. Selon les enquêtes dont on dispose, qu’il s’agisse de l’European Values Study (EVS) ou de l’International Social Survey Programme (ISSP), les individus interrogés attendent en effet du tra- vail, à des degrés divers, non seulement la possibilité de bien gagner leur vie 32 mais ils plébiscitent également l’intérêt du travail et les relations que permet celui-ci, c’est-à-dire les dimensions créatrices, expressives et relationnelles du travail. La France est à la proue de ce mouvement, les personnes interrogées depuis les années 1980 étant régulièrement les plus nombreuses en Europe (plus de 70 % selon la vague 2015 de l’ISSP) à indiquer que cette dimension est « très importante » (mais la Suisse n’est pas loin avec 61 %). Dans les travaux que nous avons consacrés à l’analyse du rapport au travail des euro- péens avec mes collègues Lucie Davoine (Davoine et Méda 2008) et Patricia Vendramin (Méda et Vendramin 2013), nous avons montré la prégnance de ces dimensions, en particulier chez les jeunes et les femmes, dimensions qui sont plus généralement liées au niveau d’éducation. Cette mise au point me semble importante parce que notre vision du futur du travail est fortement déterminée par nos représentations actuelles du travail, les catégories à travers lesquelles nous l’interprétons et les différentes formes de valorisation que nous lui accordons. Une de mes interrogations consiste à comprendre, parmi les différents discours sur le futur du travail qui saturent aujourd’hui l’espace public ou ce qui en tient lieu, quels sont ceux qui sont les plus congruents ou compatibles avec les attentes, les aspirations ou les habitus qui caractérisent aujourd’hui les rapports des européens au travail. Si nous pensons avec Bourdieu que nos aspirations sont construites, que « les agents sociaux ont des espérances ajustées à leurs chances objectives » (Bourdieu 2017, 192) et que notre libido est socialement constituée – « nous considérons des choses socialement constituées comme aimables dans un état donné d’un univers social déterminé » (Bourdieu 2017, 206) –, alors il est intéressant de voir si et comment les différents scénarios actuellement en vogue et en lice pour imaginer ou prédire le futur du travail tiennent compte de ces aspirations à un travail créateur, expressif et plus généralement – comme on l’entend sans relâche dans les entretiens – à « un travail qui a du sens » et si ces aspirations sont susceptibles de varier avec le contexte, notamment le choc qu’a constitué la pandémie de Covid-19. Cette notion de scénario est une facilité que je me donne pour regrouper des discours et des visions ayant des caractéristiques proches. La distinction stricte des scénarios – j’en propose trois – est également à vocation heuristique car à l’évidence dans le monde réel les configurations s’entremêlent. Je propose donc de considérer tois scénarios (ou configurations) du travail futur, les deux premiers étant les plus en vogue, c’est-à-dire les plus intensément promus par de nombreuses institutions, organisations internationales, think tanks, groupements d’intérêt, responsables politiques, le troisième étant quant à lui rendu nécessaire par la gravité de la situation écologique. Je m’interroge à la fois sur les conceptions du travail sur lesquelles reposent ces scénarios, sur la 33 congruence de celles-ci avec les « espérances » des citoyens des sociétés dites développées et sur les conséquences que l’on peut attendre de leur mise en œuvre sur les conditions concrètes d’exercice du travail. Je cherche également à comprendre dans quelle mesure les tendances antérieures à la pandémie ont été accélérées par celle-ci. 2.3 Avant la pandémie, trois scénarios sur le futur du travail. Le démantèlement du droit du travail (1) J’ai construit le premier scénario en rassemblant l’ensemble des recom- mandations et prescriptions inlassablement produites par les organisations internationales (OCDE, FMI, Banque Mondiale) depuis les années 1980. Il s’agit sans doute moins d’un « scénario » que d’une recommandation générale, portée par les institutions internationales qui ont également été le support de la diffusion du néo-libéralisme. Je l’appelle le scénario du « démantèlement du droit du travail ». Il s’appuie sur une conception abstraite, marchande et instrumentale du travail – ce qui correspond à la toute première couche de signification conçue par le XVIIIe siècle et magistralement explicitée par Adam Smith. Le travail est un facteur de production, dont la fonction pre- mière doit être de créer de la richesse pour l’individu (salaire), l’entrepreneur ou l’actionnaire (CA/VA) et la société (PIB). Il s’agit le plus souvent, dans l’esprit des promoteurs de cette vision, de négliger le fait que le travail occupe une place centrale dans la vie des personnes et de se débarrasser de la couche de signification arrivée le plus tardivement et de son fatras de règles venues prétendument alourdir le travail – dont l’essence reste un mélange d’effort et de création. Dans l’esprit des concepteurs de ce scénario– dont l’illustration la plus parfaite se trouve selon moi dans un très grand nombre des rapports produits par l’OCDE depuis le milieu des années 1980 (Méda 2018 ; Fretel 2016) – il nous faut nous débarrasser des règles qui font obstacle au bon fonctionnement du « marché du travail ». Le travail est en effet une marchandise qui doit pouvoir s’échanger librement sur un marché et faire l’objet des mêmes ajustements que toutes les autres marchandises. S’y opposer c’est entraîner inévitablement l’augmentation du chômage car les entrepreneur·euse·s renoncent alors à embaucher, craignant de ne pas pouvoir se séparer de leurs salariés ou d’être confrontés à des règles trop contraignant.es. C’est conformément à ces visions et prescriptions qu’ont été initiées ces trente dernières années de nombreuses réformes visant à diminuer la protection de l’emploi, à faciliter le licenciement, à faciliter les embauches 34 en contrat court, en intérim, à alléger les cotisations sociales, à remplacer la réglementation par des taxes. Les réformes allemandes, italiennes et espagnoles des années 2000, celles intervenues en France jusqu’aux Ordonnances qui ont immédiatement suivi l’élection d’Emmanuel Macron ont toutes visé à désépaissir le travail, à le ramener à sa prétendue essence, à réindividualiser les relations de travail en se débarrassant du collectif et à en revenir en quelque sorte à ce que critiquait le pourtant libéral Eugène Buret en 1840 lorsqu’il reprochait aux Français d’avoir adopté la catastrophique théorie anglaise du « travail marchandise »1. Dans cette conception, l’essence du travail, c’est le temps – et le temps, c’est de l’argent. Il n’est rien dit de l’intérêt du travail ni de ses conditions d’exercice. Si ces réformes n’entrainent en aucune manière de façon automatique des créations d’emploi bien au contraire2 (bien que ce soit pourtant leur but affiché), elles dégradent en revanche de manière cer- taine les conditions de travail comme l’ont par exemple montré les travaux que nous avons consacrés avec des collègues aux effets d’un contrat allégé (le contrat Nouvelles Embauches mis en place en 2007) sur les relations de travail ou comme l’ont montré les travaux évaluant les effets des réformes dans d’autres pays européens (Adascalitei et Pignatti Morano 2015 ; Fretel 2016 ; Vincent 2016 ; Rehfeldt 2016). Ces réformes ont accru l’insécurité pour les travailleur·euse·s, augmenté les formes atypiques d’emploi, réduit la protection, accru le nombre de contrats courts, entraîné une forte modéra- tion salariale. Le contenu du travail, la qualité de ses conditions d’exercice, la possibilité pour les individus de s’exprimer ou d’être reconnus dans leur travail n’a en rien été au cœur de ces réformes3. Ce scénario ne semble donc pas susceptible de satisfaire les immenses attentes placées sur le travail. 1 Dans son très beau texte, De la misère des classes laborieuses en France et en Angleterre, Eugène Buret dénonce la théorie du travail marchandise des Anglais qu’il considère à l’origine de la misère dans laquelle sont plongés les travailleurs des deux pays. 2 Dans leur document de travail, « Labour market reforms since the crisis: drivers and consequences » (2015), Dragos Adascalitei et Clemente Pignatti Morano analysent les déterminants et les effets des réformes de la législation sur la pro- tection de l’emploi prenant en considération 111 pays développés et en voie de développement, entre 2008 et 2014. 3 Le changement de doctrine opéré au niveau européen au moment du rapport Kok Jobs, jobs, jobs (2004) lorsqu’est abandonnée la préoccupation pour la qualité de l’emploi au profit prétendument de sa quantité illustre parfaitement cette évolution. 35 2.4 Le scénario de la Révolution technologique (2) J’en viens maintenant au deuxième scénario qui partage certains points com- muns avec le premier. Je l’ai intitulé le scénario de la Révolution technologique. C’est le plus en vogue aujourd’hui. On ne compte plus le nombre d’écrits qui s’intéressent à cette question et proposent de véritables prédictions ou prophé- ties. S’il est assez récent – le premier ouvrage scientifique et mondialement diffusé est celui de Brynjolfsson et McAfee, Race Against Machine publié en 2011 – ce scénario est en même temps assez banal dans son genre : à échéances régulières, on nous annonce en effet une vague d’innovations technologiques qui va révolutionner entièrement le travail, supprimer tous les emplois et/ ou nous délivrer de la pénibilité du travail. Cette fois-ci encore, pas question d’échapper à cette double prophétie : les discours concernant les effets de la révolution technologique sont à la fois extrêmement anxiogènes – on nous annonce d’innombrables suppressions d’emploi et même tout simplement la fin du travail (Brynjolfsson et McAfee (2011) pensent que Jeremy Rifkin avait raison lorsqu’il annonçait la fin du travail) et enthousiasmants car ce serait la fin du « travail-tripalium » : le travail devrait en effet radicalement changer de nature et devenir presque exclusivement créatif et collaboratif. Certes, il n’y a pas d’homogénéité dans ce discours : sur la question du nombre d’emplois par exemple, les chiffrages sont radicalement divergents puisque la proportion d’emplois risquant d’être automatisés varie selon les études de 47 pour cent à 9 pour cent. Ces travaux présentent néanmoins deux points communs : leur détermi- nisme technologique radical d’abord. Aucun ne considère qu’une « innova- tion » pourrait ne pas s’implanter et se diffuser du fait de la résistance des consommateur·rice·s ou des travailleur·euse·s. La question des travailleurs est réglée définitivement : on rappelle jusqu’à plus soif que les Luddites ont perdu, et qu’il ne sert donc à rien de s’opposer à ce qui est qualifié de « progrès technique ». Celle des consommateur·rice·s aussi car ils ne peuvent qu’être les principaux bénéficiaires de ces progrès, comme Taylor (2009) le soutenait déjà lorsqu’il écrivait en 1911 : « les droits du peuple sont plus importants que ceux des employeurs et des salariés ». Le deuxième point commun concerne la représentation du travail qui permet de fonder le calcul des emplois sup- primés. C’est une vision economico-ingénieuriale, un mixte de la vision de Smith et de Taylor : le travail est infiniment divisible, il est décomposable en tâches simples ou doit pouvoir l’être. D’une part, toutes les tâches simples et routinières ou routinisables seront, nous dit-on, automatisées—donc les emplois correspondants supprimés—d’autre part, le travail doit pouvoir être mis le plus possible sous la forme de tâches simples et routinisables : c’est 36 ce que les machines doivent aider les humains à faire, ou plutôt ce que les humains doivent aider les machines à faire comme le montre Antonio Casilli (2019). Mais il y a là une première simplification que certains auteurs ont remarquée et dénoncée : on ne peut pas déduire de l’automatisation d’une tâche celle de l’ensemble du « métier » ou de la profession. Non seulement parce qu’un métier est composé de plusieurs tâches, qui ne sont pas toutes également automatisables mais aussi parce que la manière dont ce métier est organisé, exercé, encastré dans une organisation varie considérablement. Valenduc et Vendramin (2019) rappellent par exemple qu’entre deux études les prévisions de suppression d’emplois pour la profession de comptable varient de 18 pour cent à 77 pour cent. Mais pourtant les discours et les efforts pour automatiser restent intenses. Car cette vision mécanique du travail s’accommode parfaitement de la représentation du marché du travail comme lieu de rencontres d’individus acheteur·euse·s et vendeur·euse·s de prestations de travail, au-delà donc des notions désormais désuètes (selon ces auteurs) de salariat et même d’entre- prise. La tâcheronisation et la plateformisation du travail sont complètement congruentes avec l’idée d’un marché du travail où les ajustements se font de manière instantanée sans besoin d’institutions pour coordonner ou réguler. Paradoxalement cette vision s’accommode aussi d’un discours enchanté sur le travail futur qui serait collaboratif, créatif, relationnel (pour ceux qui auront pu en conserver). Quelles seront donc les tâches conservées par les humains ? Des tâches hautement qualifiées selon les Susskind (2017) qui envisagent assez tranquillement la fin des professions. Des tâches qui résisteront à la routinisation, selon Frey et Osborne (2013), c’est à dire : celles qui exigent de la dextérité ; celles qui demandent une intelligence créative ; enfin celles qui nécessitent de la perspicacité sociale, de la préoccupation pour les autres. D’où l’obsession de tous les théoriciens de la Révolution technologique pour les soft skills, toutes ces compétences relationnelles, émotionnelles, prétendu- ment les seules à ne pas pouvoir être mises sous forme d’algorithme. Leur mot d’ordre est donc : innover et investir dans le capital humain, pour permettre aux individus de gagner la course contre la machine. Je voudrais ici m’arrêter un instant sur un point sans doute encore à éclaircir. Il me semble que la conceptualisation du travail sur laquelle repose toute cette vision va en fait plus loin que la conjonction de celle de Smith et de Taylor. Car il s’agit de penser le travail à la fois comme un ensemble de tâches, donc d’opérations susceptibles d’être processualisées, automa- tisées, confiées à une machine et en même temps comme un ensemble de « compétences » résistant elles à l’automatisation. Un texte extraordinaire de Foucault me semble pouvoir nous aider. Il s’agit de l’un de ses cours, la leçon 37 du 14 mars 1979, prononcé au Collège de France et publié en 2004 sous le titre Naissance de la biopolitique. Dans ce texte, Foucault défend l’idée que les économistes classiques ne se sont pas intéressés au travail : ils ont considéré que les travailleurs vendaient leur force de travail, point final. Les néo-libéraux en revanche, Schultz (1961) et Becker (1962) notamment, sont allés, soutient Foucault, beaucoup plus loin et se sont intéressés au travail lui-même, au travail comme capital permettant de fournir des intérêts (le salaire), donc comme « capital humain », ce dernier étant à la fois une source de richesse (pour l’individu et la société) et une réalité susceptible d’être profondément transformée en étant l’objet d’investissements permettant d’augmenter sa « productivité ». Un capital humain très plastique donc. Foucault a cette phrase extraordinaire : « décomposé du point de vue du travailleur en termes économiques, le travail comporte un capital, c’est-à-dire possède une aptitude, une compétence ; comme ils disent : c’est une machine » (Foucault 2004, 230) et plus loin : « il faut considérer l’ensemble comme un complexe machine/ flux, disent les néo-économistes » (Foucault 2004, 231). J’en reste là pour l’instant en m’interrogeant sur la figure de cet individu qui devient finalement le réservoir à la fois des processus automatisables (c’est une extériorisation de certaines de ses compétences) et de compétences génératrices de travail créateur, de plus en plus réduit à la portion congrue. L’individu est ainsi le support de sa propre disparition comme ressource productrice de revenus. Il est comme une machine support du transfert, de la transformation de ses propres compétences en opérations réalisées à l’extérieur de lui-même par une machine. Il est une machine interne branché sur une machine externe à laquelle il transfère ses propres compétences. Et il ne peut gagner la course contre lui-même qu’en transformant son capital humain en compétences toujours plus résistantes à la routinisation. Donc d’une certaine manière le travail humain se décompose, d’une part, en travail routinier pris en charge par les machines et, d’autre part, en travail créatif, autonome, relationnel, non formatable mais toujours en risque de l’être. Nous devons pourtant démythifier (démystifier) cette vision idyllique selon laquelle, d’un côté tous les travaux pénibles seraient pris en charge par les machines, tandis que de l’autre, le travail restant serait purement expressif, créatif, relationnel, épanouissant, autonome. La réalité est pour l’instant toute autre. D’une part, une grande partie des travaux pénibles n’est pas assez rentable pour faire l’objet d’une automatisation. Et un très grand nombre de métiers loués pour leur dimension relationnelle prétendument antinomique avec l’automatisation, comme les métiers d’aides-soignant·e·s, d’aide à domicile et plus généralement la plus grande partie des métiers du care, présentent les rémunérations parmi les moins élevées. D’autre part, 38 l’automatisation peut entraîner une dégradation et une déshumanisation du travail humain comme l’ont montré de nombreux auteurs (Casilli 2019 ; Head 2014 ; Huws 2014). Les humains ne disparaissent pas mais deviennent dépendants et contrôlés par des algorithmes qui rendent souvent leur travail encore plus insupportable (voir par exemple Rosenblat 2018). Quant à l’idée que les travailleurs seraient autonomes sur les plateformes, qu’il s’agisse de Deliveroo, d’Uber ou d’autres, de très nombreuses recherches dont celles que nous avons publiées avec mes collègues (Abdelnour et Méda 2019) mettent en évidence qu’il s’agit d’un mythe : leur autonomie est très faible, leur travail est systématiquement sous rémunéré (Berg et al. 2018), ce dont témoignent tant leurs mobilisations que les arrêts des tribunaux qui sont en train de requalifier ces travailleur·euse·s en salariés. Certes, toutes les plateformes ne se transforment pas en dispositifs d’assujettissement : les personnes qui travaillent sur la plateforme Foule Factory décrite par mes collègues Pauline Barraud de Lagerie, Julien Gros et Luc Sigalos Santos (2019) ou celles qui proposent des créations personnelles sur le site ETSY étudiées par mes collègues Anne Jourdain et Sidonie Naulin (2019) témoignent d’une autre expérience. Elles manifestent a contrario, par leur présence constante sur ce site alors qu’elles sont très faiblement rémunérées, qu’elles trouvent dans l’exercice de ce travail une possibilité de reconnaissance. Les travaux de Juliet Schor et de son équipe (2017) mettent en évidence la même variété. Finalement toute une série d’arguments doit nous rendre très méfiants vis à vis de ce scénario de la Révolution technologique et de sa capacité à satisfaire les attentes placées sur le travail. D’abord, trouver du sens à son travail ne signifie pas nécessairement éviter toute pénibilité ou toute complexité, bien au contraire. Dans l’une de mes enquêtes sur le travail (Bigi et al. 2015), j’ai interviewé une femme qui faisait toute la journée des copier-coller de numéros de chèque, qui s’exclamait : « C’est pas du travail ça ! Pour moi du travail, on doit réfléchir ! ». Je pense aussi à une thèse récente sur les néo- paysans (Wojda 2019) qui reconnaissent à la fois que leur travail est pénible mais qui disent y trouver énormément de sens. De surcroît, nous devons nous méfier du caractère performatif de ces discours qui accompagnent une prise de pouvoir et une captation de valeur et de données par quelques grandes multinationales (Zuboff 2019). Et prendre garde au fait que ce scénario fait une impasse complète sur son coût en énergie et en terres rares (Flipo et al. 2013 ; Pitron 2018), sur les dégâts de la croissance et plus généralement sur la nécessité pour nos sociétés – si elles veulent maintenir des conditions de vie authentiquement humaines sur terre selon l’expression de Hans Jonas (1985) – de s’engager au plus vite dans la Reconversion écologique. 39 2.5 Le scenario de la reconversion écologique (3) Tel est l’intitulé de mon troisième scénario, celui qui découlerait de la prise de conscience de la gravité de la situation actuelle et consisterait dans le déploiement d’une politique permettant de limiter l’augmentation du réchauffement de l’atmosphère à 1,5°C par rapport à l’ère pré-industrielle (Méda 2013 ; Cassiers et al. 2017). C’est celui que les êtres humains les plus informés et les plus raisonnables appellent de leurs vœux. Mais en quoi ce scénario serait-il plus compatible avec les espoirs que nous plaçons dans le travail ? Plusieurs études montrent qu’il devrait s’accompagner de créations d’emplois et d’un volume de travail humain plus important dans la mesure où l’usage des dispositifs et produits gourmands en énergie et/ou toxiques devrait diminuer (ILO 2018 ; Montt et al. 2018). Mais cela ne suffit pas. On peut alors imaginer, en s’appuyant sur des expériences réelles – de manière à dessiner des utopies concrètes, à la manière du regretté Erik Olin Wright –, que ces changements radicaux (que nous résumons avec mes collègues sous le terme de « société de post-croissance ») soient organisés de manière à per- mettre le déploiement d’un nouveau régime de travail. Adepte de la théorie de l’effondrement, l’ancien Ministre écologiste français de l’environnement, Yves Cochet, a publié un ouvrage (Cochet et al. 2019) consacré à la des- cription de ce que sera le monde après l’effondrement. Une population très réduite – par les maladies, les guerres et la mort – vivra dans des bio-régions : une grande partie de la population sera redevenue agricole et subviendra à ses besoins. On peut en effet imaginer une forme d’anti-déversement (le contraire de ce que décrivait l’économiste français Alfred Sauvy dans sa théorie du déversement) avec une forte augmentation de la population dans un secteur agricole moins productiviste, et dans un secteur secondaire organisé bien différemment d’aujourd’hui non plus autour de grandes multinationales, de cascades de sous-traitants et de longues chaînes de valeurs mondiales mais autour de coopératives ou de petites organisations plutôt artisanales, au terme d’un processus de relocalisation des productions. André Gorz imaginait un artisanat high tech (Gorz 2008). Trebor Scholz (2013) propose le déploie- ment d’un coopérativisme de plateformes qui commence à se concrétiser et de nombreuses expérimentations concrètes commencent à dessiner la possi- bilité d’un changement de paradigme. Un tel projet de nouveau régime de travail a été conçu de façon très précise par Simone Weil – une philosophe française profondément et charnellement engagée dans l’étude du travail et de ses changements désirables, dans l’Enracinement : il est très congruent avec l’idée de reconversion écologique car il est fondé sur la limitation des besoins. Mettre un coup d’arrêt à la dérive moderne qu’a connue le travail 40 soutient Simone Weil, c’est d’abord parvenir à proportionner les besoins et la consommation pour que le travail reste raisonnable et ne soit pas emporté par la recherche incessante des gains de productivité. C’est donc « chercher l’organisation la plus humaine compatible avec un rendement donné ». Com- ment faire ? Il s’agit d’abord de revoir la conception même des machines (en opposition radicale avec certaines manières actuelles d’automatiser le travail) : « Jusqu’ici les techniciens n’ont jamais eu autre chose en vue que les besoins de la fabrication. S’ils se mettaient à avoir toujours présents à l’esprit les besoins de ceux qui fabriquent, la technique entière de la production devrait être peu à peu transformée » (Weil 1949, 46). Ensuite, il s’agit de développer des organismes industriels d’une espèce nouvelle, de petits ateliers organisés sur le mode coopératif et de renoncer à la grande industrie. Quant aux sociétés anonymes, il n’y aurait peut-être pas d’incon- vénient, en ménageant un système de transition, à les abolir et à les déclarer interdites […]. En tous cas, un tel mode de vie sociale ne serait ni capitaliste ni socialiste. Il abolirait la condition prolétarienne, au lieu de vouloir l’étendre à tous. (Weil 1949, 57) Il ne s’agit finalement pas pour elle de rendre le travail joyeux ou épanouis- sant – ce qui apparaît soudainement comme un nouveau mythe – mais de l’améliorer en mettant au centre des objectifs la dignité. Dans la conférence prononcée le 10 septembre dernier lors du congrès de la société suisse de sociologie qui est à l’origine du présent texte, je m’interrogeais sur le possible déclencheur du changement en me demandant pourquoi et par quoi celui-ci pourrait être provoqué : « Sera-ce le résultat d’une entreprise raisonnable, conséquence d’une prise de conscience du coût écologique de notre mode de développement ou du coût social et politique des mauvais emplois, comme l’ont mis en évidence de nombreux travaux ? Je pense notamment aux recherches qui montrent le lien entre mauvais emplois et vote pour les partis d’extrême-droite et au plaidoyer récent de Rodrick et Sabel (2019) en faveur d’une économie des bons emplois. Mais aussi à toutes les mobilisations des travailleurs contre les grandes plateformes et aux mobilisations pour l’écologie. Sera-ce au contraire le résultat d’un choc externe, semblable à celui que les sociétés kabyles pré-capitalistes décrites par Bourdieu ont subi, – des pénuries, des guerres, des épidémies accompagnant le début de l’effondrement ? Dans la version optimiste des choses, on peut penser qu’il sera possible d’observer un réencastrement polanyien, de la société dans la nature, de l’économie dans la société, du travail dans la société. Le travail redeviendrait peut-être alors ‹ l’autre nom de l’activité économique qui accompagne la vie elle-même › » (Polanyi 1983), et pourrait retrouver, au-delà du capitalisme, sa juste place » 41 indiquais-je. J’évoquais pour terminer l’ouvrage majeur de Moishe Postone, Temps, travail et domination sociale (2010) et notamment cette note dans laquelle l’auteur s’interroge sur ce que sera le travail après le capitalisme. Postone soutient que le travail cessera de jouer un rôle socialement médiatisant et que sa place sera finalement réduite. Il précise ce que sera la nature de ce travail lorsque la loi de la valeur aura été abolie : Le fait que Marx pense que, dans la société future, le travail social sera structuré de manière à être satisfaisant et agréable ne signifie pas qu’il pense que ce travail deviendra un jeu. L’idée que Marx se fait du travail non aliéné, c’est que celui-ci est exempt de domination sociale directe ou abstraite. Il peut du même coup devenir une activité de réalisation de soi et partant se rapprocher du jeu. Toutefois cette liberté par rapport à la domination ne signifie pas une liberté par rapport à toutes les contraintes. (Postone 2010, 58‒59) Je concluais ainsi la conférence : « De nombreuses questions restent néanmoins en suspens autour de cette vision : notamment le rôle joué par la technologie et celui joué par le revenu universel. Les ouvrages récents qui témoignent de la révolte des individus contre les conditions du travail modernes et l’absence de sens du travail dans le capitalisme, de même que les expérimentations concrètes visant à vivre et travailler autrement, en lien avec la crise écologique, sont peut-être le signe que ce changement est en cours. Souhaitons que les idées que nous développons dans ce congrès contribuent, comme les changements structurels – économiques et écologiques – en cours, à l’émergence ce que Bourdieu appelait un nouveau cosmos ». Mais quelques mois après le congrès, la pandémie de Covid-19 frappait nos sociétés. 2.6 Trois scénarios à l’épreuve de la crise de la Covid-19 La crise sanitaire a accéléré certaines des tendances décrites ci-dessus, constituant un choc de très grande ampleur sur les travailleur·euse·s et les organisations. Elle a d’abord révélé et exacerbé les inégalités au sein du monde du travail. En effet, alors que les populations de nombreux pays étaient confinées, certains travailleurs étant pris en charge par des dispositifs de chômage partiel, d’autres exerçant leurs activités en télétravail, un grand nombre de travailleur·euse·s continuaient à assurer les activités essentielles, permet- tant à leurs concitoyen·ne·s de survivre : dans des villes devenues vides, les soignant·e·s, les personnels de vente, les vigiles, les personnels des métiers 42 du care, les chauffeur·euse·s de bus et de taxi, les personnels de nettoyage et d’entretien, les livreur·euse·s occupaient l’espace, devenant soudainement visibles alors même qu’ils appartenaient jusqu’alors pour la plupart à des catégories peu médiatisées, voire peu considérées (à l’exception des médecins et des infirmie·re·s). Des enquêtes ont très rapidement mis en évidence  – par exemple les remarquables enquêtes réalisées par l’Office for National Statistics (ONS) britannique –, que ces « travailleur·euse·s essentiel.le.s » étaient non seule- ment les plus exposés à la contamination par le virus, mais avaient également payé un très fort tribut à la pandémie. L’une de ces études a ainsi analysé les 2 494 décès impliquant le coronavirus, intervenus entre le 9 mars et le 20 avril dans la population en âge de travailler (vingt à soixante-quatre ans) en Angleterre et au Pays de Galles (Office for National Statistics 2020). La profession étant indiquée sur le certificat de décès, il est possible de comparer la composition socioprofessionnelle des personnes décédées du Covid-19 à celle de l’ensemble des personnes décédées du même âge et du même sexe. Les plus forts taux de surmortalité concernent en premier lieu les travailleuses des métiers du soin à la personne (hors travailleurs de la santé), suivis des vigiles, des chauffeurs de taxi et d’autobus, des chefs cuisiniers et des assistants de vente et de détail et plus généralement des travailleurs peu qualifiés. L’ONS a aussi montré la plus forte probabilité pour les non-Blancs de décéder du coronavirus, en partie explicable par des facteurs socio-économiques. Ces études – qui ne peuvent pas pour l’instant être réalisées en France, car les instituts statistiques n’y ont pas légalement l’autorisation de relier origine ethnique, cause médicale de décès et profession – permettent de démontrer la plus grande vulnérabilité de certaines professions et auraient du inciter à mieux les protéger (notamment les personnes atteintes par ailleurs de mala- dies chroniques), en matière d’équipements – qui ont cruellement manqué en début de crise –, mais aussi de statut d’emploi et de conditions de travail. En effet, les emplois des key workers sont aussi, constate l’ONS, ceux qui sont les moins bien payés, qui présentent les conditions de travail les plus difficiles et les statuts les plus précaires. Ces conditions socio-économiques sont aussi en cause dans la prévalence élevée de comorbidités (obésité, diabète, hyper- tension), dont la présence accroît le risque de décès en cas de Covid-19. De la même manière, des chercheuses ont pu mettre en évidence qu’en France, les classes populaires payaient un très lourd tribut à la Covid-19 (Khlat et Counil 2020), en raison de la conjonction de conditions socio-économiques peu favorables (faibles revenus, logements exigus, rationnement des soins, mauvaise alimentation) comme l’a mis en évidence l’enquête Coconed de 43 l’Ined (Lambert et al. 2020) et de conditions d’emploi se caractérisant par de fréquents contacts peu protégés et des difficultés à exercer un droit de retrait. Si de nombreuses promesses ont été faites dans plusieurs pays pour améliorer la rémunération et les conditions de travail des travailleur·euse·s essentiel·le·s, les mesures d’urgence prises au plus fort de la pandémie ont souvent consisté à remettre en cause un certain nombre de droits des travailleurs et à ignorer le dialogue social alors que la sortie de la phase de confinement s’accompa- gnait de propos insistants visant à demander aux salariés de travailler plus et d’accepter une modération salariale. Les mois qui ont suivi la première vague de la crise sanitaire ont vu s’intensifier massivement les licenciements et le nombre de personnes au chômage. Dès lors, on peut craindre que la préoccupation d’amélioration des conditions de travail et d’emploi ne soit pas centrale. Il sera important de comprendre si les mêmes erreurs que celles faites après la crise de 2008 se renouvellent : comme nous l’avons mis en évidence (Heyer et al. 2018), un certain nombre de mesures composant une politique de retour à l’équilibre des finances publiques et de désinflation compétitive ont été mises en œuvre en Europe à partir de 2011, qui ont contribué à étouffer la reprise et à prolonger la crise dans la plupart des pays européens. Au contraire, l’acceptation de forts déficits, l’amélioration de la qualité de l’emploi et l’augmentation de salaires des travailleurs essentiels devrait constituer une des lignes de force du plan de relance sur lequel les pays européens sont enfin parvenus à s’accorder. Il faut en effet noter que si les mauvaises conditions d’emploi ont accéléré la contamination d’un certain nombre de travailleurs, a contrario, la qualité de l’emploi – contrats perma- nents, salaires décents, santé et sécurité – constitue un élément constitutif essentiel de la résilience des travailleurs. L’investissement dans la qualité de l’emploi comme dans la santé apparaît dès lors comme un élément essen- tiel dans la prévention des dégâts de la pandémie. Mais celle-ci a entraîné d’autres bouleversements majeurs : en enjoignant à une partie importante des travailleur·euse·s de continuer leurs activités à partir de leur domicile en télétravail, elle a sans nul doute fait franchir à cette modalité d’exercice du travail une étape essentielle, de même qu’à la digitalisation de l’économie. En effet, de nombreux travailleurs mais aussi leurs employeurs ont soudain découvert les possibilités insoupçonnées du télétravail pendant que les achats en ligne explosaient par impossibilité ou peur du contact. Les possibles conséquences de cette expérience sont à n’en pas douter immenses. Pour ne prendre que l’exemple français, avant la crise sanitaire, le télétravail était peu répandu : seuls 3 pour cent des salariés déclaraient le pratiquer au moins un jour par semaine, dont plus de 60 pour cent de cadres. La crise sanitaire a fait exploser ce chiffre : un quart des salariés était considérés en télétravail à 44 la fin mars 2020. Or, avant la crise, 80 pour cent des personnes interrogées déclaraient être favorables à la mise en œuvre d’un plan de déploiement du télétravail en France, 59 pour cent souhaitant elles-mêmes télétravailler et beaucoup considérant la méfiance des employeur·euse·s comme la principale raison du faible développement de cette forme de travail. De nombreux travailleur·euse·s ayant goûté la formule et des employeur·euse·s s’étant converti·e·s ou ayant été convaincu·e·s de l’intérêt de la développer (des économies de coût de bureaux pouvant être faites), la tendance à la diffusion du télétravail pourrait bien s’affirmer fortement. Notons cependant que si, à petite dose, le télétravail comporte, pour ceux qui peuvent le pratiquer, de nombreux avantages (il permet aux salariés d’effectuer moins de déplace- ments – ce qui est aussi bon pour le climat – et de gagner du temps et de la concentration. Il peut améliorer la conciliation entre travail et famille dans certains cas), il comporte aussi de très nombreux inconvénients, devenus plus évidents avec sa diffusion récente. Il prive les salariés de certaines dimensions du travail pourtant essentielles : le contact physique, les échanges informels, les interactions sont des éléments constitutifs du travail qui ne peuvent dura- blement disparaître qu’au prix d’une dégradation des conditions d’exercice de l’activité. Même si des logiciels performants ont permis l’organisation de réunions, la prise de décision, la poursuite du travail, les salarié·e·s ont aussi fait l’expé- rience du caractère gravement insuffisant des interactions par écran interposé, de la fatigue engendrée par ce type d’échange, de la baisse de concentration qu’elle engendre, mais surtout des risques inhérents à l’isolement. Le.la télétravailleur·euse est isolé·e, privé·e du soutien du collectif de travail, seul·e face à une éventuelle surcharge de travail ou à des consignes inaccessibles, incapable de voir comment réagissent les collègues, de bénéficier de leur aide, de se mobiliser éventuellement. Par ailleurs, l’expérience du télétravail a également mis en évidence le considérable brouillage entre vie personnelle et vie professionnelle engendré par le télétravail : la présence de la famille, notamment de jeunes enfants, mais aussi d’autres adultes, rend le travail plus difficile. L’exiguïté des logements, la biactivité, le nombre insuffisant d’équi- pements constituent autant de phénomènes aggravants qui ne font souvent qu’entraîner la reproduction des inégalités à l’œuvre dans la société : on sait déjà que le déséquilibre dans la prise en charge des activités domestiques et familiales entre les hommes et les femmes s’est accru durant le confinement (Lambert et al. 2020). La pandémie a également démultiplié les achats en ligne, le recours à la livraison (qu’il s’agisse de biens ou de services) et s’est accompagnée d’une explosion du recours à des applications numériques fortement soumises à la 45 critique auparavant. On peut craindre que les habitudes prises pendant la pandémie ne persistent et n’entraînent de profondes reconfigurations dans l’organisation du travail en accroissant tout à la fois le nombre de dispositifs de contrôle et de surveillance, mais plus généralement l’individualisation et la tâcheronisation du travail, ainsi qu’une nouvelle segmentation au sein de la société entre le travail au contact (délégué en tant que « sale boulot » aux travailleurs les moins qualifiés) et travail protégé. Le travail pourrait donc s’effectuer de moins en moins au sein de collectifs de travail dans des lieux partagés, en co-présence, et de plus en plus de façon individualisée, la coor- dination étant assurée par des algorithmes, comme cela est déjà le cas dans de nombreux secteurs. Les tendances à l’œuvre avant la crise, externalisation, re-taylorisation, individualisation, délégation du « sale boulot » (Hugues 1996), tâcheronisation et plateformisation du travail (Casilli 2019 ; Abdelnour et Méda 2019 ; Rosenblat 2018) pourraient donc s’aggraver avec la pandémie. On risque fort de voir se développer le recours aux applications numériques de livraison à la demande, qu’il importe de réguler si nous voulons éviter des phénomènes tels que ceux décrits dans le dernier film de Ken Loach « Sorry we missed you » ou dans une littérature de plus en plus abondante, et les entrepôts logistiques où les conditions de travail ne sont pas meilleures (Gabo- rieau 2015), en plus de l’extension de la gestion du travail par le numérique (Jounin 2017) qui s’apparente trop souvent à une privation d’autonomie pour les travailleur·euse·s et à une déshumanisation du travail (Bernard 2020). Comme avant la pandémie, la solution pour enrayer ce processus délétère pour les travailleur·euse·s consiste, d’une part, à réinstaurer de forts méca- nismes de régulation permettant d’éviter que les GAFA ne bénéficient de privilèges exorbitants – donc notamment de les empêcher de recourir à des travailleur·euse·s indépendants dont le statut est « fictif » et de les obliger à respecter leurs devoirs d’employeur·euse·s (notamment en payant des cotisa- tions sociales et en assurant la sécurité des travailleur·euse·s) – et, d’autre part, à exiger de la part de toutes les organisations le respect de normes sociales et environnementales strictes. Car contrairement à la digitalisation sans frein qui porte en germe le risque d’augmenter les émissions de CO2 en plus de déshumaniser le travail, la reconversion écologique est plus que jamais une voie de choix pour le monde du travail. La sortie de la crise sanitaire exige en effet des plans de relance massifs qui ne peuvent se permettre d’accélérer le changement climatique, à un moment où les prévisions n’ont jamais été aussi inquiétantes. Une relance verte massive qui permettrait à nos sociétés de rebâtir complètement leurs fondements en devenant décarbonées pourrait permettre la création d’emplois nombreux et utiles, non délocalisables et sans doute beaucoup plus susceptibles 46 de satisfaire le besoin de sens qu’un grand nombre d’emplois actuels. On peut en effet penser que les emplois visant à faire reculer les manifestations majeures du changement climatique et à reconvertir nos sociétés auraient plus de sens et seraient considérés comme plus utiles que ceux qui contribuent aujourd’hui à la destruction de notre planète. C’est le sens du Manifeste « Démocratiser, Démarchandiser, Dépol- luer » que nous avons initié avec Isabelle Ferreras et Julie Battilana (2020), porté avec neuf autres collègues femmes, qui a été signés par plus de 6 000 scientifiques dans le monde entier et que nous avons publié sous le titre : Le Manifeste Travail. Démocratiser, Démarchandiser, Dépolluer. Il appelle à mettre en œuvre une profonde démocratisation des entreprises et plus généralement de toutes les organisations de travail en exigeant la parité entre investisseurs en travail et investisseurs en capital dans la gestion et la direc- tion des entreprises. Il propose également de rendre effective une garantie de l’emploi particulièrement indispensable pour aider les travailleur·euse·s à surmonter la crise sanitaire mais aussi pour permettre à nos sociétés engagées dans la reconversion écologique d’anticiper et d’assurer les transferts massifs d’emploi qui devront s’opérer des secteurs très émetteurs de CO2 aux secteurs et aux métiers des économies décarbonées. La conviction sous-jacente des auteures est que même s’ils sont attachés à leur emploi et n’accepteront sans doute pas de gaieté de cœur la fermeture de leurs usines ou la suppression de leurs emplois, les travailleur·euse·s sont les plus à même de co-décider les modalités des reconversions et d’imaginer la manière dont celles-ci doivent s’opérer au mieux. S’il n’est pas capable à lui seul de garantir la réussite d’un processus émi- nemment périlleux, ce dispositif semble néanmoins pouvoir en constituer une pièce maîtresse. La participation des travailleur·euse·s à sa conception et à sa gestion n’est évidemment pas sans rapport avec le surcroît de sens que ceux-ci devraient normalement en retirer. Mais ce ne sont pas seulement de puissants dispositifs d’accompagnement des transitions – faisant de celles-ci des transitions justes – dont nous aurons besoin mais également de processus de planification et de coordination pré- cis permettant d’organiser la répartition territoriale des nouveaux emplois que la reconversion écologique devrait créer si nous parvenons à engager les moyens financiers nécessaires. Pour ce qui concerne la France, deux études récentes, l’une produite par l’ADEME et figurant en annexe du rapport du Haut Conseil pour le Climat (2020), l’autre par le WWF (2020), ont chiffré les possibles créations d’emplois à l’horizon 2030 à 600 000 pour la première et deux millions pour la seconde, en particulier dans le bâtiment et les transports. Un groupe a également travaillé à produire un Plan de 47 programmation des emplois et des compétences qui a tenté de chiffrer les créations et suppressions d’emploi par secteur mais devrait être prolongé par des travaux plus fins qui devront articuler les différentes échelles, nationale et territoriales. C’est d’une planification semblable à celle qui a permis à la France d’organiser sa reconstruction après la Seconde Guerre Mondiale dont nous avons aujourd’hui besoin dans l’ensemble des pays européens et au niveau de l’Union européenne elle-même. 2.7 Conclusion En repartant de l’histoire longue du travail, j’ai tenté de distinguer dans quelle mesure différents scénarios dessinant l’avenir du travail étaient en mesure de satisfaire les immenses attentes qui s’expriment à l’égard du travail. J’ai essayé de mettre en évidence qu’avant la pandémie de Covid-19, ni le scénario bien engagé en Europe de démantèlement du droit du travail, ni celui en vogue de Révolution technologique n’étaient susceptibles de faire droit à ces attentes, contrairement au scénario de la Reconversion écologique dont le déploiement apparaît le plus nécessaire. La crise du Covid-19 est venue bouleverser ou plutôt accélérer les tendances à l’œuvre antérieurement en rendant plus pro- bable le développement de la digitalisation de nos économies mais encore plus désirable et nécessaire la reconversion écologique de nos sociétés. Une partie du dénouement est aujourd’hui dans les mains des instances européennes, Commission et Parlement, qui sont en capacité de faire de notre continent le plus avancé dans la lutte contre la crise écologique et de transformer en profondeur les conditions d’exercice du travail. 2.8 Références bibliographiques Abdelnour, Sarah et Méda Dominique. 2019. Les nouveaux travailleurs des applis. Paris: Presses universitaires de France. Abdelnour, Sarah et Sophie Bernard. 2019. Quelles résistances collectives face au capitalisme de plateforme ? In Sarah Abdelnour et Dominique Méda (dir.), Les nouveaux travailleurs des applis. Paris: Presses universitaires de France, 61‒74. Adascalitei, Dragos et Clemente Pignatti Morano. 2015. Labour Market Reforms since the Crisis: Drivers and Consequences. Research Department Working Pa- per No. 5. Geneva: International Labour Office. October, 2015, https://www. ilo.org/wcmsp5/groups/public/---dgreports/---inst/documents/publication/ wcms_414588.pdf (28.02.2021) 48 Arntz, Mélanie, Terry Gregory et Ulrich Zierahn. 2016. The Risk of Automa- tion for Jobs in OECD Countries: A Comparative Analysis. OECD Social, Employment and Migration Working Papers No. 189. Paris: OECD Publishing. 14 May, 2016, https://www.oecd-ilibrary.org/docserver/5jlz9h56dvq7-en.pdf ?expires=1611933101&id=id&accname=oid009212&checksum=F961E73C 210612777749C69E7D784524 (28.02.2021) Barraud de Lagerie, Pauline et Luc Sigalos Santos. 2019. Les plateformes de micro- travail : le tâcheronnat à l’ère numérique. In Sarah Abdelnour et Dominique Méda (dir.), Les nouveaux travailleurs des applis. Paris: Presses universitaires de France, 33‒46. Becker, Gary 1962. Investment in human capital: a theoretical analysis. Journal of Political Economy, 70(5). Berg, Janine, Marianne Furrer, Ellie Harmon, Uma Rani et Six Silberman. 2018. Digital Labour Platforms and the Future of Work: Towards Decent Work in the Online World. Geneva: ILO. 2018, https://www.ilo.org/ wcmsp5/groups/public/---dgreports/---dcomm/---publ/documents/publication/ wcms_645337.pdf (28.02.2021) Bernard, Sophie. 2020. Le nouvel esprit du salariat. Paris: Presses universitaires de France. Bigi, Maëlle, Olivier Cousin, Dominique Méda, Laetitia Sibaud et Michel Wie- viorka. 2015. Travailler au XXIème siècle. Des travailleurs en quête de reconnais- sance. Paris: Robert Laffont. Bourdieu, Pierre, Claude Seibel, Alain Darbel et Jean-Paul Rivet. 1963. Travail et travailleurs en Algérie. Paris: Mouton. Bourdieu, Pierre. 2017. Anthropologie économique. Cours au collège de France. Paris: Raisons d’Agir/Seuil Brynjolfsson, Erik et McAfee Andrew. 2011. Race Against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy. Lexington, MA: Digital Frontier Press. Casilli, Antonio. 2019. En attendant les robots. Paris: Le Seuil. Cassiers, Isabelle, Kevin Maréchal et Dominique Méda (dir.). 2017. Post-growth Economics and Society: Exploring the Paths of a Social and Ecological Transition. Abingdon et New York: Routledge. Chamley, Pierre. 1965. La doctrine économique de Hegel et la conception hégé- lienne du travail, in Nicolin F. et Otto Pöggeler, Hegel-Studien Band 3 Cochet, Yves, Agnès Sinaï, et Benoît Thévard. 2019. Biorégions 2050. Paris: Forum Vies Mobiles. Davoine, Lucie et Dominique Méda. 2008. Place et importance du travail en Europe : une singularité française ? Document de travail No. 96, Centre d’Études de l’Emploi. https://hal.archives-ouvertes.fr/hal-00276220 (28.02.2021) 49 Durkheim, Emile. 2013. De la division du travail social. Paris: Presses Universi- taires de France Ferreras, Isabelle, Julie Battilana, et Dominique Méda. 2020. Le Manifeste Travail. Démocratiser, Démarchandiser, Dépolluer. Paris: Le Seuil. Flipo, Fabrice, Michelle Dobré, et Marion Michot. 2013. La face cachée du numé- rique. L’impact environnemental des nouvelles technologies. Montreuil: L’Échappée. Foucault, Michel. 2004. Naissance de la biopolitique, Cours au Collège de France. Paris: Gallimard. Fretel, Anne. 2016. Réformes du marché du travail : que disent les « exemples » étrangers ? Chronique Internationale de l’IRES, 3(155): 3‒25. Frey, Carl Benedikt et Osborne, Michaël. 2013. The Future of Employment: How Susceptible are Jobs to Computerization. Oxford Martin Programme on Technol- ogy and Employment. Oxford: Université d’Oxford, https://www.oxfordmartin. ox.ac.uk/downloads/academic/future-of-employment.pdf (28.02.2021) Freyssenet, Michel. 1999. Invention, centralité et fin du travail. Paris: CSU. Gaborieau, David. 2015. Travailler sous commande vocale dans les entrepôts de la grande distribution. In Thébaud-Mony, Annie, Pierre Davezies, Laurent Vogel, et Serge Volkoff (dir.), Les risques du travail. Pour ne pas perdre sa vie à la gagner. Paris: La Découverte, 219‒222. Gorz, André. 2008. Ecologia. Paris: Galilée. Haut Conseil pour le Climat. 2020. Redresser le Cap, Relancer la transition. Rapport Annuel 2020. Paris : Haut Conseil Pour le Climat. Juillet, 2020, http://www.observatoireclimat-hautsdefrance.org/Les-ressources/Ressources- documentaires/Rapport-annuel-2020-Haut-Conseil-pour-le-Climat-Redresser- le-cap-relancer-la-transition(28.02.2021) Head, Simon. 2014. Mindless: Why Smarter Machines are Making Dumber Humans. New York, NY: Basic Books. Heyer, Eric, Pascal Lokiec, et Dominique Méda. 2018. Une autre voie est possible. Paris: Flammarion. Hugues, Edward. 1996. Le regard sociologique. Essais choisis. Textes rassemblés et présentés par Jean-Michel Chapoulie. Paris: Éditions de l’EHESS Huws, Ursula. 2014. Labor in the Global Digital Economy: The Cybertariat Comes of Age. NYU Press. Inglehart, Robert. 1990. Culture Shift in Advanced Industrial Society. Princeton, NY.: Princeton University Press. Jonas, Hans. 1985. The Imperative of Responsibility. In Search of an Ethics for the Technological Age. Chicago: University of Chicago Press. Jounin, Nicolas. 2017. Le caché de la Poste. La genèse de temps virtuels pour organiser le travail des facteurs. La Revue de l’Ires, 3(93): 25‒50. 50 Jourdain, Anne et Sidonie Naulin. 2019. Marchandiser ses loisirs sur Internet : une extension du domaine du travail ? In Sarah Abdelnour et Dominique Méda (dir.), Les nouveaux travailleurs des applis. Paris: Presses universitaires de France, 47‒60. Khlat, Myriam et Emilie Counil. 2020. Covid-19 : les classes populaires paient- elles le plus lourd tribut au coronavirus en France ? Paris: The Conversation. https://hal.archives-ouvertes.fr/hal-02938233 (28.02.2021) Kok, Wim. 2004. Jobs, Jobs, Jobs. Creating More Employment in Europe. Report of the Employment Taskforce Chaired by Wim Kok. Luxembourg: Office for Official Publications of the European Communities. Lambert, Anne, Joanie Cayouette-Remblière, Elie Guéraut, Guillaume Le Roux, Catherine Bonvalet, Violaine Girard, et Laetitia Langlois. 2020. Le travail et ses aménagements : ce que la pandémie de covid-19 a changé pour les Français. Population et sociétés, 579(7): 1‒4. Marx, Karl. 1979. Œuvres, tome II: Economie, Collection Bibliothèque de la Pléiade. Paris: Gallimard. Méda, Dominique. 1998. Le travail, une valeur en voie de disparition? Paris: Champs- Flammarion. Méda, Dominique. 2013. La mystique de la croissance : Comment s’en libérer ? Paris: Flammarion. Méda, Dominique. 2018. Code du travail et chômage. Genèse d’une mise en accusation. Droit ouvrier, Mars, 236: 132-139. Méda, Dominique et Patricia Vendramin. 2013. Réinventer le travail. Paris: Presses universitaires de France. Montt, Guillermo, Kirsten Svenja Wiebe, Marek Harsdorff, Moanas Simas, Antoine Bonnet, et Richard Wood. 2018. L’action pour le climat, une action contre l’emploi ? Évaluation des conséquences du scénario à 2° C sur l’emploi. Revue Internationale du Travail, 157(4): 573‒613. National Office of Statistics. 2020, Coronavirus (Covid-19) Related Deaths by Occupation, England and Wales: Deaths Registered up to and Including 20  April 2020. 11  May, 2020, https://www.ons.gov.uk/peoplepopulation- andcommunity/healthandsocialcare/causesofdeath/bulletins/coronavirus- covid19relateddeathsbyoccupationenglandandwales/deathsregistereduptoand- including20april2020 (28.02.2021) Organisation internationale du Travail. 2018. Emploi et questions sociales dans le monde 2018: Une économie verte et créatrice d’emploi. Genève : OIT. Pitron, Guillaume. 2018. La guerre des métaux rares: La face cachée de la transition énergétique et numérique. Paris: Les liens qui libèrent. Polanyi, Karl. 1983. La grande transformation. Paris: Gallimard. Postone, Moshe. 2010. Temps, travail et domination sociale. Paris: Les Editions de Minuit. 51 Rehfeldt, Udo. 2016. Le « Jobs Act », un nouveau pas vers la flexibilité pour les travailleurs et vers la sécurité pour les employeurs. Chronique Internationale de l’Ires, 155(3): 57‒70. Rifkin, Jeremy. 1995. The End of Work: The Decline of the Global Labor Force and the Dawn of the Post-Market Era. New York: G.P. Putnam’s Sons. Rodrik, Dani et Charles F. Sabel. 2020. Building a Good Jobs Economy. HKS Working Paper No. RWP20-001, Harvard Kennedy School. January, 2020, https://www.hks.harvard.edu/publications/building-good-jobs-economy# citation (28.02.2021) Rosenblat, Alex. 2018. Uberland—How Algorithms are Rewriting the Rules of Work. Berkeley: University of California Press. Scholz, Trebor (dir.). 2013. Digital labor. The Internet as Playground and Factory. New York et London: Routledge. Schor, Juliet. 2017. Does the Sharing Economy Increase Inequality within the Eighty Percent?: Findings from a Qualitative Study of Platform Providers. Cambridge Journal of Regions, Economy and Society, 10(2): 263–279. Schultz, Theodore W. 1961. Investment in human capital. American Economic Review, vol. 51: 1–17. Smith, Adam. 1999. La richesse des nations. Paris: GF. Susskind, Daniel et Richard Susskind. 2016. The Future of the Professions. How Technology Will Transform the Work of Human Experts. Oxford University Press. Taylor, Fredrick 2009. The Principles of Scientific Management. Cornell University Library. Valenduc, Gérard et Patricia Vendramin. 2019. La fin du travail n’est pas pour demain. Notes de prospective, 6(6): 1‒16. Vernant, Jean-Pierre. 1996. Mythe et pensée chez les Grecs. Paris: La Découverte. Vincent, Catherine. 2016. Dans la crise, des réformes du marché du travail radi- cales mais au bilan incertain. Chronique Internationale de l’Ires, 3(155): 44‒56. Weil, Simone. 1949. L’enracinement. Prélude à une déclaration des devoirs envers l’être humain. Paris: Gallimard. Wojda, Loïc. 2019. L’émancipation paysanne : une redéfinition du travail, de la consommation et de la richesse. Penser les reconversions professionnelles vers l’agri- culture paysanne avec Pierre Rahbi et au-delà. PhD dissertation, Faculté des Sciences Économiques et Sociales, Université de Lille, France. Wright, Erik Olin. 2010. Envisioning Real Utopias. London: Verso. WWF France. 2020. Monde d’après : l’emploi au ceour d’une relance verte. Juillet 2020, https://www.wwf.fr/sites/default/files/doc-2020-07/20200710_ Rapport_Monde-apres-emploi-au-coeur-relance-verte_WWF-min.pdf (28.02.2021) Zuboff, Shoshana. 2019. The Age of Surveillance Capitalism. The Fight for a Human Future at the New Frontier of Power. New York: Public Affairs. 52 3 Trajectories of Change at Work: A Global Look at Technological Change and Informalization Chris Tilly 3.1 Introduction Two momentous trends have dominated the discussion of global workplace change over the last two decades. The first is massive increases in computing power that have facilitated the collection and analysis of big data, and hence the addition of a digital overlay to a growing range of activities—what is sometimes called the ‘fourth industrial revolution’ (Schwab 2016, passim). Uber’s platform-based ride-hailing service exemplifies this metamorphosis, though it just represents the tip of a much larger iceberg. The second is a degradation of labour standards and escalation of worker insecurity and vulnerability in large sections of employment, a process that some have called precaritization and others have called informalization. Though Uber can once more be adduced as an example, perhaps more striking is the case of semi-skilled manufacturing workers in many countries, from the United States and Japan to China, India, and Mexico, who have tumbled from the labour aristocracy to the precariat (Doellgast et al. 2018; Tilly et al. 2013). Both trends continue long trajectories of change at work that have fas- cinated social researchers at least as far back as Marx. My aim in this paper is to use the current iterations of these transformations to demonstrate and argue for an analytical framework that emphasizes the embeddedness of work in broader social, economic, and political processes, the centrality of power in determining directions of change, and the critical role of learning in both types of change. This framework is actor-centred, spotlighting the varied roles of key actors in the world of work: firms, workers, and the state. The paper continues by briefly elaborating the proposed framework. Then I discuss labour informalization and precaritization through the three lenses of embeddedness, power, and learning. Following, I examine work- 53 place technological change through the same three lenses. The goal is not to attempt a comprehensive nor even representative review of the large and rapidly growing literature about how work is evolving on these two fronts, but to illustrate the application of the framework. I draw on examples from global South and North alike; again my purpose is not to make systematic comparisons but to spotlight an analytical approach that helps to understand shifts in both broad settings. 3.2 Embeddedness, Power, Learning Consider in more detail the three central concepts of embeddedness, pow- er, and learning. I use embeddedness in a somewhat more expansive way than Granovetter (1985). In Granovetter’s sense, embeddedness means that economic processes take place within a matrix of social relations. In this chapter, embeddedness refers to immersion of economic processes within social relations and larger-scale economic and political structures. This broader formulation helps distinguish among processes and mechanisms that occur at different scales. I define power in the usual Weberian sense of an actor’s ability to impose their will over others (Weber 1978). As Weber and sociologists since have emphasized, power takes many forms. Within the workplace, an employer may exercise despotic or hegemonic power, in Michael Burawoy’s (1983) terms. Firms and trade unions can use market power, but also coercive power (Bowles et al. 2013)—in the case of unions, these are the two varieties of structural power analysed by Wright (2000). Both types of actors can also use what Wright called associational power, power built by coordinating advocacy by numerous individuals or organizations, to sway the state. For states, power is closely linked to capacity: a state body may be able to compel an individual capitalist to comply with a directive, but may lack the capacity to apply the directive more generally. In considering learning, I focus on learning by organizations, be they firms (or networks or associations of firms), worker groups, or state entities. Though learning of well-established patterns and routines makes up an im- portant part of this terrain, particularly of interest are cases of learning as a process of experimentation to solve complex, previously unsolved problems with uncertain outcomes (Sabel and Reddy 2014). Stories about embeddedness, power, and learning are populated by actors. At the most abstract level, the central actors in these workplace accounts are employers, workers, and the state. But to more fully account for how these 54 processes take place, it is necessary to aggregate up and disaggregate down. Aggregating up means examining industry associations, worker organizations, and state coalitions (for example, in trade agreements), as well as less formal networks of these three types of actors. Disaggregating down recognizes that a complex organization is never a single actor: firms and states have multiple subunits, factions, and internal networks, all of which can and do engage in contention, alliance-building, and consent. With this conceptual apparatus in place, we can now turn first to infor- malization and precaritization, and then to technological change, to trace out the tracks of embeddedness, power, and learning at work. 3.3 Informalization and Precaritization Informalization and precaritization are overlapping but distinct concepts. Informal employment is compensated work that, while legal, is not governed by the same regulations, legal standards, and social security arrangements that characterize other work (Castells and Portes 1989). Informalization, then, involves shifting work outside the reach or grasp of employment law—where the law’s reach refers to the types of work and workers that are nominally covered by the law, and its grasp refers to effective enforcement of extant laws (Zatz 2008); the concept can also refer to relaxation or reduction of legal standards that makes formal work more similar to informal employment. Precarious labour is employment that falls short of the ‘standard employment relationship’ (SER) available to core workers, which may include provisions well beyond basic legal protections, such as fringe benefits and expectations about long-term job security (Vosko 2006). Precaritization thus, analogous to informalization, can refer either to shifting labour outside of the SER, or downgrading of the standards that constitute the SER. Analysts focusing on the global South have tended to privilege the concept of informality, whereas those studying the North have more often utilized the concept of precarity, though increasingly use of both ideas has filtered across regions (Mosoetsa et al. 2016). Following Lee et al. (2020), I identify five main theories of informalization of work. Two, dualism and legalism, are by now relatively discredited. The dualist approach suggested that informal labour was a vestigial remnant of traditional forms of work and enterprise and would disappear with modern- ization, but the renewed growth of informal employment in regions such as Latin American and India undermined that premise. Legalist perspectives, exemplified by De Soto’s (1989) work, suggest that informalization is the result 55 of imposition of excessive restrictions on businesses, but regulatory flexibili- zation has not tamped down informality, and has in some cases increased in step with it. Three other theories remain viable. A survivalist analysis points to the growth of vulnerable workforces with few alternatives, such as migrants, as a driver of informalization. Structuralist theories attribute informalization to capital’s restructuring of employment. And disembedding theories suggest that ‘the West is following the Rest’, in Jan Breman’s (2013, 130) words: after several decades of regulations defending labour standards, wealthy countries are reverting to the historical and global norm of work that is regulated minimally or in ways repressive rather than supportive of workers. To some extent, as Lee and co-authors (2020) have suggested, survivalist, structuralist, and disembedding analyses are consistent, differing mainly in whether they centre the worker, the employer, or the state as an actor. In explaining informality and precarity, it may be tempting to privilege embeddedness and power over learning, and embeddedness in social relations and institutions over embeddedness in larger economic structures. After all, isn’t the spread of degraded work primarily the result of the capital-labour power balance tilting more toward capital, and the consequent rolling back of protective institutions? I answer this question ‘yes’, but also argue that learning and economic embeddedness play central roles in these shifts. 3.3.1 Embeddedness and Job Degradation Let us start by stipulating that firms, employment sectors, and labour markets are embedded in a scaffolding of laws and institutions, primarily established at the national level (though with substantial sub-national components). Analysts have advanced numerous taxonomies of such legal-institutional structures (Gallie 2007); one useful one is Itzigsohn’s (2000) two-dimensional grid that arrays state economic policies from developmental to predatory on one axis, and the labour regulatory regime from protective to repressive on the other. Itzigsohn argues that more predatory and repressive institutions foster more informality. Correspondingly we can suppose that movement in those directions will facilitate informalization. Examples include regimes that have reduced basic labour standards, such as Chile after the 1973 coup (Cook 2007) and China in the economic reform period that began in the late 1970s (Andreas 2019). These two countries have also made institutional changes that undermined worker organizations’ strengths at various points: for example, limiting unions to the enterprise level in Chile, and in China restricting the independence and funding sources of 56 the ‘labour NGOs’ that constitute workers’ main autonomous voice (Chan et al. 2016). Another set of laws specify the boundaries of different types of work. In the United States, a series of court cases and state laws in the 1950s‒1960s established that workers hired through a temporary help agency are solely employees of the agency, not the contracting employers—opening the door to growth of this new precarious category (Gonos 1997). In an ongoing con- flict, in 2019 that country’s largest state, California, passed a law specifying a strict test for the ‘independent contractor’ status that Uber, Mechanical Turk, and other platform-based companies rely on, attempting to limit growth of that category (McNicholas and Poydock 2019), but then in 2020 Uber and other platform-based transportation companies secured voter approval for a ballot proposition that grants them an exemption from the law, spending $ 200 million in their successful campaign (Conger 2020). States can, on the one hand, crack down on informal enterprises and employment or, on the other hand, fail to enforce laws setting labour stan- dards. Governments in Mexico and the United States have shut down day labour assembly points, and street vendors (Carré and Tilly 2017; Sarmiento and Tilly 2016), but have simultaneously turned a blind eye to widespread violations of labour law in other sectors (Bernhardt et al. 2008; Cook 2007). Governments can also feed downward pressure on job quality by facilitat- ing the flow of migrants labouring on unequal terms, as with Europe’s laws allowing ‘posting’ of migrant workers subject to the (typically lesser) labour protections of their home countries (Berntsen and Lillie 2015). Firms are not just embedded in institutional frameworks, but also in sectors and economies. Precaritization and informalization are shaped by enterprises’ immersion in markets for good and services, labour, and capital, as well as by available technologies. Examples from retail in the United States illustrate. In the market for goods and services, once some retailers lowered prices to consumers by expanding part-time employment at a lower pay rate and with scant benefits, other retailers felt compelled to follow suit to avoid being priced out of the market (Tilly 1996). In the labour market, the warehouse store chain Costco pays an above-market wage to reduce labour turnover and boost worker motivation, but if large numbers of other retailers tried to do the same the higher wage would no longer be above-market, and the payoffs reaped by Costco would evaporate (Carré and Tilly 2020). In the capital market, small mom-and-pop retailers have little access to capital, hence limited ability to upgrade their operations, and frequently stay afloat by evading labour standards (Carré and Tilly 2017). Finally, on the techno- logical front, the availability of demand forecasting and scheduling software 57 has fueled retailers’ adoption of fragmented and unpredictable work sched- ules, to match store headcount as closely as possible to consumer demand (Carré and Tilly 2020); the Instacart shopping service has hired hundreds of thousands of ‘shoppers’ uncovered by employment laws or social security, thanks to a phone-based app that allows the company to claim, Uber-style, that the workers are ‘independent contractors’ (Kang 2020). 3.3.2 Power and Job Degradation The over-arching story of informalization and precaritization centres on power. This is a familiar argument, so I will not go into detail, but will sketch out some general points and a few examples. Above all, the global turn to neoliberal policies—including the transformations in Chile and China de- scribed above as well as Germany’s creation of the less-protected ‘mini-job’ and the U.S. government’s shift to a lower real minimum wage and weakening of labour laws—result from capital gaining power relative to labour. Thus, the laws and institutions in which firms and labour markets are embedded are not simply a legacy, but an actively contested battleground. Capital also wields a form of power over the state by non-compliance, which often goes hand-in-hand with a government policy of non-enforcement (Bernhardt et al. 2008; Cook 2007). Power is divided and stratified within the state as well. For example, though China’s national-level policy over the past fifteen years has prioritized formalizing work contracts and improving worker compensation to build a broad internal market for goods, local governments tend to favour businesses, which expand the local state’s licit and illicit revenue sources, over workers (Ngok 2008). There is often subnational variation in local labour policies and in rigor of enforcement of nominally universal national policies—for example, between southern and northern coastal China (Zhou 2020), be- tween Brazil’s South and Northeast (Baltar et al. 2010), or between the U.S. Northeast and west coast on the one hand, and southern and central regions on the other (Meyerson 2015). Federalism, in which regions and localities have substantial autonomy in setting policy, can at times lead to stronger labour protections filtering up from the local to the national level, as in the cases of U.S. minimum wage laws (Sonn and Luce 2008) and laws protecting domestic workers (Tilly et al. 2019). Workers muster their own collective power to resist or mitigate informal or precarious conditions. We follow Agarwala (2018) and Evans and Tilly (2016) in listing several types of worker resistance movements. Marxian movements are workers’ classic struggles against capitalists to improve wages and working 58 conditions (Silver 2003). In Polanyian movements, workers and allies press the state for labour regulations, social insurance systems, and the like (Polanyi 1944). Agarwala (2018) christens worker movements that primarily seek recognition, for example, via official classification as workers or issuance of ID cards, as Fraserian. And Evans and Tilly (2016) coin the term Owenite to describe solidarity economy-type initiatives to create a parallel economic sphere guided by humanist rather than profit considerations. Technology availability can alter the forms and potentials of worker resistance, as with Indonesian delivery drivers or Chinese Walmart workers using social media platforms to build collective power (Ford and Honan 2017; Tapia et al. 2018). Worker organizing is, of course, uneven across sectors, and those sectoral differences themselves differ across nations. Consider construction and domes- tic work, two sectors prone to informalization. In India, informal construction workers were the first to win tripartite welfare boards providing a basic social insurance system, and to expand these welfare boards from regional to national institutions, whereas other sectors including domestic work have achieved success later and less completely (Agarwala 2013; Agarwala 2018). In Mexico, construction workers initially built stronger and more durable trade unions than domestic workers, but in recent years construction unions’ willingness and ability to protect workers has waned, whereas domestic worker associa- tions have achieved greater strength and scope (Tilly and Rojas-García 2021). Capital, likewise, is no unified monolith. Associations, alliances, and factions of business exercise power in distinct ways and toward distinct, and sometimes conflicting, ends (Schulze-Cleven and Weishaupt 2015; Weil 2008). Even within a single corporation, research on U.S. retail finds conflicts between goals of customer service (promoted by Marketing departments), of employee engagement (promoted by Human Resources), and of compli- ance with labour laws (promoted by Legal as well as Human Resources), on the one hand; and on the other, the overriding goal of keeping costs low by minimizing the labour budget (a goal dictated by Finance) (Carré and Tilly 2017). Though financially dictated strategies typically exert strong political clout within firms, managers sometimes subvert those dictates, for example, building up new multi-level hierarchies within call centres that were conceived of as flat organizations (Moss et al. 2008) or treating workers hired through temporary help agencies as essentially permanent employees (Moss et al. 2005). 3.3.3 Learning and Job Degradation Though the foregoing description of power in processes of informalization and precaritization may make it appear that actors are deploying well-es- 59 tablished repertoires, in reality actors are constantly learning—developing, replicating, and diffusing new action strategies. The spread of fissuring in workplaces is precisely the story of learning and diffusion of a series of initially novel arrangements: franchising a branded business, using temp agencies, outsourcing and offshoring parts of production processes, using software platforms to hire workers as independent contractors, and so on (Weil 2017). Growing non-compliance with labour laws in wealthy countries like the United States is also learned behaviour (Bernhardt et al. 2008)—in some cases, as with union-busting, under the direct tutelage of specialist consultants (Bronfenbrenner 2009). Likewise, the erosion of standard corporate fringe benefits in the United States diffused over time as businesses imitated each other (Boushey and Tilly 2009). Like firms, states learned to make work more precarious. Chile’s neoliberal policies were developed by economists trained in the conservative University of Chicago economics department; in turn, other countries in Latin America learned from Chile (Cook 2007). The neoliberal ‘Washington consensus’ spread globally in part because multilateral lenders imposed structural adjustment plans that required this policy package, but also in part because national elites saw it as advantageous (Gore 2000). At times labour precaritization practices piloted in one part of the country then get replicated across the country, as with the liberalized labour regimes in southern China (Andreas 2019) or the maquiladora zone Mexico established along its border with the United States (Bensusán and Cook 2003). Conversely, states sometimes learn to reregulate precarious work, as in the case of labour inspection regimes in Central America and the Caribbean (Piore and Schrank 2018). Labour organizing also follows learning pathways. Along China’s coast, the formation of labour NGOs percolated from South to North, but orga- nizations also worked out distinctive forms in each region (Zhou 2020). In the United States, living wage movements spread through imitation as well as through replication and scaling up by national organizations (Sonn and Luce 2008; Tilly 2005). Cross-nationally, global union federations have in- troduced new strategies to their affiliated national unions, as well as learning from each other (Ford 2019). The same is true for global networks of informal workers; for example, India’s Self-Employed Women’s Association initiated a global network of home-based workers that has subsequently spread across multiple regions (WIEGO 2020a). Institutions such as the International La- bor Organization and, on a smaller scale, Women in Informal Employment, Globalizing and Organizing (WIEGO) have also played an important role in spreading innovative strategies (WIEGO 2020b). 60 3.4 Technological Change1 Unlike discussions of labour informalization, analyses of technological change focus above all on learning (Cascio and Montealegre 2016). But the diffusion of technologies and of particular means of implementing them, with consequences for labour, is equally structured by embeddedness and power. Technologies applied to work are, after all, adopted by people within organizations. This discussion of technological change will be briefer than the preceding discussion of informalization and precaritization, but will examine the same three dimensions of the process. I will pay particular attention to the current wave of digitally driven technological change—the ‘fourth industrial revolution’ (Schwab 2016)—and its implications for work. 3.4.1 Learning and Technological Change New technologies enter organizations in a variety of ways: through direct invention, imitation, provision by a vendor, promulgation by a broker (such as a consultant), or migration of an individual with the relevant knowledge from one organization to another. Amazon invented the technology under- lying Amazon Go stores, which implement cashier-less checkout via sensors that identify a customer by their cell phone and cameras and other sensors that track what that shopper is taking off the shelf; subsequently several other companies have imitated that technology. Uber pioneered the platform-based ‘sharing economy’ model of employment of workers as independent contrac- tors, and since then many other enterprises have attempted to establish the ‘Uber of ’ some new sector. Whereas Uber entered the ride-hailing business itself, Kronos, the pioneer and leader in employee scheduling software, plays a vendor role, selling successive generations of software to large employers (Kronos 2020). Consultants can also be energetic proselytizers for techno- logical change. Frederick Winslow Taylor promoted ‘scientific management’ as one of the earliest business consultants, with devastating effects on labour. Today, McKinsey, Forrester, and scores of other consulting firms convince businesses to adopt new digital technologies and show them how to do it. Technological innovations can also travel as individuals with the relevant know-how move from one business to another or start their own businesses. The classic example is Fairchild Semiconductor, the original trunk on Silicon Valley’s family tree, from which dozens of engineers left to establish their own startups (Laws 2017). But similar stories exist across all sectors and countries: for example, the story of Desh Garments, which used specialized knowledge 1 Where not otherwise identified, the source is Carré and Tilly (2020). 61 to introduce export-oriented garment production to Bangladesh, only to have dozens of managers leave to start their own apparel firms (Easterly 2001). As the example of Taylorism suggests, technologies—in many cases the same technology—can be used to empower workers, increase their discretion, and enrich their jobs, or alternatively to closely direct, surveil, and control them, jealously guarding the underlying knowledge as a management prerogative. What determines which way(s) a technology will be used? The answers lie in embeddedness and power. 3.4.2 Embeddedness and Technological Change Cascio and Montealegre (2016) contrast three perspectives on the relationship between technology and organizations (as well as a fourth, ‘absent presence’ perspective that is silent on that relationship). Exogenous force models posit that technology develops according to its own logic, and its availability propels organizational change. Emergent force theories view technology as thoroughly embedded in specific, historically and socially determined organizational con- texts, so that technologies’ impact emerges from the interaction between the technical and the social. Entanglement in practice suggests that technology and organizations (and roles within them) are not separate forces that interact at a given point, but two mutually constitutive aspects of sociotechnical reality than only and always exist in relationship with each other. For the purposes of this essay, I deploy an emergent force approach, while acknowledging that entanglement may provide a more satisfactory description of some techno- logical change processes. To start with, technological change, like informalization, is embedded in an institutional matrix. Consider the technical organization of work in retail stores in the United States, France, and Germany. France has a high minimum wage, restrictions on the number of retail establishments within a local area as well as limits on store opening hours, and an educational system that leaves teenaged youth with little time for part-time jobs. The result is a capital-intensive organization of work, with a relatively small number of highly paid workers managing a rapid throughput of customers. Germany’s tripartite apprenticeship system, designed for craft production, supplies retailers with apprentices and graduates of the two-year retail program, all with a broad base of skills unlike that of workers in the other two countries. Retail stores’ work organization is therefore skill-intensive, with workers multi-tasking and taking on duties that in other countries fall to managers. The United States has a low minimum wage, minimal restrictions on store construction or opening hours (so that the country as a whole is heavily ‘overstored’), a 62 less demanding educational regime that makes young workers available for low-wage work, and a weak vocational education system with little if any retail-related training. Not surprisingly, U.S. retail stores are labour-intensive, with little expectation of or investment in worker skills, and high labour turnover (Carré and Tilly 2017). The incorporation of new digital technologies in retail stores has likewise been influenced by laws and other institutions. China’s ‘anything goes’ policies on artificial intelligence and facial recognition along with other biometrics, the Chinese government’s concerted investments in skills and research to promote tech industries, and weak protections of intellectual property have helped make it the world leader in retail innovation (though market factors, such as a labour shortage and a growing middle-class seeking convenience and novelty, also matter; I discuss market embeddedness below). The European Union, on the other hand, imposes limits on facial recognition and is about to promulgate regulations on the use of artificial intelligence and algorithm-based decision-making, and new retail technologies have spread at a slower pace there (Drozdiak 2020). Retailers in the United States have lagged behind those in both East Asia and Europe, in part because the country’s weak labour protections keep labour cheap, reducing the incentive to replace labour via technology. To take a less well-known case, Mexico was very slow to adopt e-commerce, in part because until recently credit cards were rare—a fact un- derpinned by a dysfunctional postal system and underdeveloped institutions governing consumer finance (in-person instalment payments remained the main form of consumer credit for decades) (Carré and Tilly 2017). Firms’ adoption of technology is also embedded in markets. Regarding consumer markets, luxury retailers specialize in personal attention and hard discounters compete on price with little attention to service levels, so neither of these segments is a good candidate for new technologies; mass marketers aiming at the broad middle, on the other hand, invest in tech to increase choices, convenience, and service while reducing labour costs. In the United States, capital markets favour new retail startups like Amazon and dozens of newer entrants (via a well-developed venture capital market) and established giants like Walmart, Target, and Kroger; many other retailers have been tar- geted by private equity firms that sold their real estate and other assets and burdened them with debt, leaving them poorly positioned to access capital for technology investments. In the labour market, East Asian investments in technology are propelled in part by labour shortages, in China’s case due to a combination of rapid economic growth and the one-child policy, in the case of a number of other countries (notably Japan) due to rapid ageing of the 63 population. U.S. retailers became much more interested in tech investments as a labour shortage sharpened in the late 2010s. 3.4.3 Power and Technological Change Varied dimensions of power mould the trajectory of technology adoption, as Thomas (1994), among others, has pointed out forcefully. The familiar dimension of the capital-labour balance of power weighs heavily. Unionized coal miners, longshore workers, nurses, meat-cutters, professors, and others have fought to limit the use of labour-saving technology or to give workers more control over the use of technology—usually with limited success (Bore- ham et al. 2007). The ways firms choose to implement new technologies, in turn, can affect the degree of individual and collective power mustered by workers. For example, powerful, inexpensive cameras and other sensors backed up by automated analytics can closely monitor worker actions. But not all technological advances further worker subordination. The assembly line reduced skill-based structural power, but increased disruption-based structural power; extensive subcontracting linked to ‘lean’, just-in-time inven- tory systems disrupt workers’ associational power by scattering them across separate firms, but can again endow workers at key points in the supply chain with greater disruptive power (Silver 2003). The growing ubiquity and falling cost of digital, analytical, and communications technologies may open the pathway to Owenite exit by workers to form cooperatives or peer production networks (though the lack of adaptation of capital markets to these forms of enterprise still poses an obstacle) (Evans and Tilly 2016). Power struggles within the firm do not end with labour-management tussles, but also include divisions among executives, managers, and engi- neers, a point Thomas (1994) emphasizes. Highly centralized power within a firm can lead to rapid execution of a planned technological makeover, but can also, by limiting critical feedback, lead to large and rapid investments of resources into ill-conceived technological plans. Examples include General Electric’s attempt to reinvent itself as a software and data analysis company, which ended disastrously (Mann and Gryta 2020), or Amazon’s investment of hundreds of millions of dollars and nearly a decade of R&D into the Amazon Go ‘frictionless’ retail technology, which has resulted in what Business Week describes as a glorified convenience store (Stone and Day 2019). But the more typical case is the fragmentation of power. On the one hand, CEO’s can see their visions for new technology diverted or subvert- ed by the engineers on whom they depend (Thomas 1994). On the other hand, Chief Technology Officers in non-tech industries (such as retail) are 64 often outflanked by more senior, better-networked executives in finance and operations, limiting their ability to lead technological transformations. For that matter, actual implementation of technology in a large company can depend on cooperation from thousands of middle and frontline managers; their passive resistance can stall tech reboots (Carré and Tilly 2020). And of course, power within and over the state determines the laws and state-supported institutions that made an appearance in the discussion of embeddedness above. As we saw from cross-national comparisons, the exercise of power in debates over policy on consumer privacy, land use, intellectual property, and the educational system can exert far-reaching influence on whether and how firms use available technologies. Power in other aspects of policy matters as well. One obvious example is trade policy. And for that matter, the policy conundrums, power struggles, and collective action problems summarized as developing countries’ ‘middle-income trap’ create obstacles to the diffusion of up-to-date technologies and the launching of home-grown technological innovation (Doner and Schneider 2016). Indeed, closing the circle in my analysis, one common aspect of this trap is the spread of informalization that expands a section of capital with little interest in innovation or technological upgrading, contributing to a political impasse on technologically-oriented policies. 3.5 Conclusion: Some Lessons for Action Current trajectories of change at work around the world presage a continuation of informalization and precaritization and of rapid technological change. I have argued the informalization and precaritization, on the one hand, and technological change, on the other, are decisively shaped by a combination of power, embeddedness, and learning. To some extent, this argument is a polemic against narrower analyses that are common in the literatures on these two topics. The literature on labour informalization tends to emphasize either embeddedness (as in the survivalist literature, which depicts informalized labour as embedded in labour markets characterized by excess labour supply) or power (as in structuralist and disembedding theories, which attribute informalization to choices by capital and the state). I argue that both, and learning as well, are critical to an adequate understanding of informalization. Literature on technological change typically spotlights learning. I argue that embeddedness and power (in its many dimensions—power of management vs labour, power relations within management, and power of various actors vis à vis the state) are important parts of the story. 65 But ultimately, so what? What guidance do these insights yield for action regarding the future of work? Suppose we specify as goals the achievement of decent work and the direction of new technologies to meet human needs, including worker needs. Based on this, I would suggest a few points that mark the beginning of a discussion rather than anything approaching a full answer to this question. First, workers and their allies must build power, relative to both capital and the state. This will require all four modalities of organizing noted above: Marxian labour organizing directed at capital, Polanyian organizing directed at the state, Fraserian organizing for recognition from the capital and the state, and Owenite organizing to withdraw from capitalist production rela- tions. Such power-building will not only involve traditional forms such as trade unions and parties, but alternative coalitions, networks, associations, and other organizations. Second, we need to construct and strengthen institutions that will re-em- bed work, regulating and controlling capital to halt and reverse processes of informalization and precaritization. To the extent possible, regulatory mechanisms should universalize worker protections, ‘taking them out of competition’ rather than allowing businesses to compete via lower costs and prices achieved at the cost of labour standards. A revitalized pro-worker institutional infrastructure must offer opportunities to exit survivalism and degraded work, whether through a growing mass of better-regulated jobs, Polanyian decommodification of labour (as with a generous universal basic income), or Owenite construction of an alternative economy (which can be facilitated by new forms of financial regulation to support cooperatives and the like, and relaxation of intellectual property restrictions on the sharing of technology). Finally, building the necessary power and institutions will require much new learning. Building power will require experimentation with new organizational forms and strategies, an experimentation process that is well underway around the world (Evans and Tilly 2016). Designing adequate protective institutions will require learning about new forms of informalization of work (as in the proliferating varieties of workplace fissuring), new technologies that threaten worker well-being (e.g. new surveillance techniques), and sometimes both at the same time (as in the case of platform-based enterprises that informalize labour by design). Understanding change at work in terms of power, embeddedness, and learning does not answer all our questions about the future of work, let alone our questions about how to improve that future. But the analysis offers vital tools toward addressing those questions, not least by setting a research agenda. 66 Figuring out how to construct power, restore embeddedness, and accelerate the learning needed toward these two ends—this is the challenge before us. 3.6 References Agarwala, Rina. 2013. Informal Labor, Formal Politics, and Dignified Discontent in India. Cambridge, UK: Cambridge University Press, 2013. Agarwala, Rina. 2018. ‘The Development of Labor under Contemporary Capi- talism.’ Sociology of Development 4(3): 239–60. Andreas, Joel. 2012. ‘Industrial Restructuring and Class Transformation in China.’ In Beatriz Carrillo and David Goodman (Eds.), China’s Peasants and Workers: Changing Class Identities. Northampton, MA: Edward Elgar. Andreas, Joel. 2019. Disenfranchised: The Rise and Fall of Industrial Citizenship in China. Oxford, UK: Oxford University Press. Baltar, Paulo, Anselmo dos Santos, José Dari Krein, Eugenia Leone, Marcelo Proni, Amilton Moretto, Alexandre Maia, and Carlos Salas. 2010. Moving toward Decent Work—Labour in the Lula Government: Reflections on Recent Brazilian Experience. Working Paper No. 9. GLU Working Papers. Berlin: Global Labour University. http://www.global-labour-university.org/fileadmin/ GLU_Working_Papers/GLU_WP_No.9.pdf (accessed March 14, 2021). Bensusán, Graciela and Maria L. Cook. 2003. ‘Political Transition and Labor Revitalization in Mexico.’ In Daniel Cornfield and Holly McCammon (Eds.), Labor Revitalization: Global Perspectives and New Initiatives. Amsterdam, Netherlands: Elsevier. Bernhardt, Annette, Heather Boushey, Laura Dresser, and Chris Tilly (Eds.). 2008. The Gloves-Off Economy: Labor Standards at the Bottom of the American Labor Market. Ithaca, NY: Cornell University Press. Berntsen, Lisa, and Nathan Lillie. 2015. Social Dumping at Work: Uses and Abuses of the Posted Work Framework in the EU. ETUI Policy Brief. Brussels, Belgium: European Trade Union Institute. https://www.etui.org/sites/default/files/ Policy%20Brief%202015.07%20Berntsen%20Lillie.pdf (accessed March 14, 2021). Boreham, Paul, Rachel Parker, Paul Thompson, and Richard Hall. 2007. New Technology @ Work. London, UK: Routledge. Boushey, Heather and Chris Tilly. 2009. The Limits of Work-Based Social Support in the United States. Challenge, 51(2): 81‒114. Bowles, Samuel, Frank Roosevelt, Richard Edwards, and Mehrene Larudee. Un- derstanding Capitalism: Competition, Command, and Change. Fourth edition. Oxford, UK: Oxford University Press, 2017. 67 Bronfenbrenner, Kate. 2009. No Holds Barred: The Intensification of Employer Opposition to Organizing. Briefing Paper. Washington, D.C.: Economic Policy Institute. 20 May, 2009, https://www.epi.org/publication/bp235/ (accessed March 14, 2021). Breman, Jan. 2013. A Bogus Concept? New Left Review 84: 130–38. Burawoy, Michael. 1983. Between the Labor Process and the State: The Changing Face of Factory Regimes Under Advanced Capitalism. American Sociological Review 48(5): 587–605.. Carré, Françoise and Chris Tilly. 2017. Where Bad Jobs Are Better: Retail Jobs Across Countries and Companies. New York, NY: Russell Sage Foundation. Carré, Françoise and Chris Tilly, with Chris Benner and Sarah Mason. 2020. Change and Uncertainty, Not Apocalypse: Technological Change and Store-Based Retail. Berkeley: UC Berkeley Labor Center and Working Partnerships USA. Cascio, Wayne F. and Ramiro Montealegre. 2016. How Technology Is Changing Work and Organizations. Annual Review of Organizational Psychology and Organizational Behavior, 3(1): 349‒375. Castells, Manuel, and Alejandro Portes. 1989. ‘World Underneath: The Ori- gins, Dynamics, and Effects of the Informal Economy.’ In Manuel Castells, Alejandro Portes, and Lauren Benton (Eds.), The Informal Economy: Studies in Advanced and Less Developed Countries. Baltimore and London: The Johns Hopkins University Press. Chan, Anita, Kaxton Siu, and Sarah Swider. 2016. Interpreting Chinese Labour: Informalisation or Empowerment? Made in China 4: 23–21. Conger, Kate. 2020. Uber and Lyft Drivers in California Will Remain Contractors. New York Times, November 4, 2020. https://www.nytimes.com/2020/11/04/ technology/california-uber-lyft-prop-22.html (accessed March 14, 2021). Cook, Maria Lorena. 2007. The Politics of Labor Reform in Latin America: Between Flexibility and Rights. University Park, PA: Pennsylvania State University Press. De Soto, Hernando. 1989, The Other Path: The Invisible Revolution in the Third World. New York, NY: Harper & Row. Doellgast, Virginia, Nathan Lillie, and Valeria Pulignano, eds. Reconstructing Solidarity: Labour Unions, Precarious Work, and the Politics of Institutional Change in Europe. Oxford, UK: Oxford University Press, 2018. Doner, Richard F. and Ben Ross Schneider. 2016. The Middle-Income Trap: More Politics than Economics. World Politics, 68(4): 608‒644. Drozdiak, Natalia. 2020. Even the Pandemic Doesn’t Stop Europe’s Push to Regulate AI. Bloomberg Business Week. 7 April, 2020, https://www.bloomberg. com/news/articles/2020-04-07/coronavirus-isn-t-stopping-europe-s-push-to- regulate-ai (accessed May 10, 2020) Easterly, William. 2001. The Elusive Quest for Growth. Cambridge, MA: MIT Press. 68 Evans, Peter and Chris Tilly. 2016. ‘The Future of Work: Escaping the Current Dystopian Trajectory and Building Better Alternatives.’ In Stephen Edgell, Heidi Gottfried, and Edward Granter (Eds.), Sage Handbook of the Sociology of Work and Employment. Thousand Oaks, CA: Sage Publications, 651‒671. Ford, Michele. 2019. From Migrant to Worker: Global Unions and Temporary Labor Migration in Asia. Ithaca, NY: Cornell University Press. Ford, Michele, and Vivian Honan. 2017. ‘The Go-Jek Effect.’ In Edwin Jurriens and Ross Tapsell (Eds.) Digital Indonesia: Connectivity and Divergence. Sin- gapore: ISEAS. Gallie, Duncan, Ed. 2007. Employment Regimes and the Quality of Work. Oxford, UK: Oxford University Press. Gonos, George. 1997. The Contest over ‘Employer’ Status in the Postwar United States: The Case of Temporary Help Firms. Law and Society Review 31(1): 81–110. Gore, Charles. 2000. The Rise and Fall of the Washington Consensus as a Paradigm for Developing Countries. World Development, 28(5): 789‒804. Granovetter, Mark 1985. Economic Action and Social Structure: The Problem of Embeddedness. American Journal of Sociology, 91(3): 481‒510. Itzigsohn, José. 2000. Developing Poverty: The State, Labor Market Deregulation, and the Informal Economy in Costa Rica and the Dominican Republic. University Park, PA: Pennsylvania State University Press. Kang, Jaewon. 2020. “Grocers Embrace Food Delivery, but They Still Don’t Love It.” Wall Street Journal, June 13, 2020. https://www.wsj.com/articles/grocers- embrace-food-delivery-but-they-still-dont-love-it-11592056800 (accessed March 14, s2021). Kronos. 2020. Kronos History: An inspiring history of innovation, https://www. kronos.ca/about-us/kronos-history (accessed August 24, 2020). Laws, David. 2017. Fairchild Semiconductor: The 60th Anniversary of a Silicon Valley Legend. Computer History Museum. CHM Blog—Curatorial Insights (blog), 19  September, https://computerhistory.org/blog/fairchild-semicon- ductor-the-60th-anniversary-of-a-silicon-valley-legend/ (accessed October 21, 2020). Lee, Byoung-Hoon, Sarah Swider, and Chris Tilly. 2020. Informality in action: A relational look at informal work,” International Journal of Comparative Sociology 61(2-3): 91-100. Mann, Ted and Thomas Gryta. 2020. The Dimming of GE’s Bold Digital Dreams. Wall Street Journal. 18 July, 2020, https://www.wsj.com/articles/the-dimming- of-ges-bold-digital-dreams-11595044802 (accessed July 19, 2020). McNicholas, Celine, and Margaret Poydock. 2019. “How California’s AB5 Protects Workers from Misclassification.” Fact Sheet. Economic Policy Institute, No- vember 2019. https://files.epi.org/pdf/179103.pdf (accessed March 14, 2021). 69 Meyerson, Harold. 2015. How the American South Drives the Low-Wage Economy. The American Prospect, Summer 2015. https://prospect.org/labor/american- south-drives-low-wage-economy/ (accessed March 14, 2021). Mosoetsa, Sarah, Joel Stillerman, and Chris Tilly. 2016. Precarious Labor, South and North: An Introduction. International Labor and Working-Class History 89: 5–19. Moss, Philip, Hal Salzman, and Chris Tilly. 2008. Under Construction: The Continuing Evolution of Job Structures in Call Centers. Industrial Relations, 47(2): 173‒208. Moss, Philip, Hal Salzman, and Chris Tilly. 2005. ‘When Firms Restructure: Understanding Work-Life Outcomes.’ In Ellen Kossek and Susan Lambert (Eds.), Work and Life Integration in Organizations: New Directions for Theory and Practice. Mahwah, NJ: Lawrence Erlbaum Associates. Ngok, Kinglun. 2008. The Changes of Chinese Labor Policy and Labor Legislation in the Context of Market Transition. International Labor and Working-Class History 73: 45–64. Piore, Michael J. and Andrew Schrank. 2018. Root-Cause Regulation: Protecting Work and Workers in the Twenty-First Century. Cambridge, MA: Harvard University Press. Polanyi, Karl. 1944. The Great Transformation. Boston: Beacon Press. Sabel, Charles, and Sanjay Reddy. 2014, Learning to Learn: Undoing the Gordian Knot of Development Today. Challenge 50(5): 73–92. Sarmiento, Hugo and Chris Tilly. 2018. Governance Lessons from Urban Infor- mality. Politics and Governance, 16(1): 199-202. Schulze-Cleven, Tobias and J. Timo Weishaupt. 2015. Playing Normative Lega- cies: Partisanship and Employment Policies in Crisis-Ridden Europe. Politics & Society, 43(2): 269‒299. Schwab, Klaus. 2016. The Fourth Industrial Revolution. New York: Crown Business. Silver, Beverly J. 2003. Forces of Labor: Workers’ Movements and Globalization Since 1870. Cambridge, UK: Cambridge University Press. Sonn, Paul and Stephanie Luce. 2008. ‘New Directions for the Living Wage Movement.’ In Annette Bernhardt, Heather Boushey, Laura Dresser, and Chris Tilly (Eds.), The Gloves-Off Economy: Labor Standards at the Bottom of the American Labor Market. Ithaca, NY: Cornell University Press. Stone, Brad and Matt Day. 2019. Amazon’s Most Ambitious Research Project Is a Convenience Store. Bloomberg Business Week. 18  July, 2020, https:// www.bloomberg.com/news/features/2019-07-18/amazon-s-most-ambitious- research-project-is-a-convenience-store (accessed August 14, 2020). Tapia, Maite, Manfred Elfström, and Denisse Roca-Servat. 2018.Bridging Social Movement and Industrial Relations Theory: An Analysis of Worker Organiz- ing Campaigns in the United States and China. Research in the Sociology of Organizations 56: 173–206. 70 Thomas, Robert. 1994. What Machines Can’t Do: Politics and Technology in the Industrial Enterprise. Berkeley, CA: University of California Press. Tilly, Chris. 1996. Half a Job: Bad and Good Part Time Jobs in a Changing Labor Market. Philadelphia, PA: Temple University Press. Tilly, Chris. 2005. ‘Living Wage Laws in the United States: The Dynamics of a Growing Movement.’ In Maria Kousis and Charles Tilly (Eds.), Economic and Political Contention in Comparative Perspective. Boulder, CO: Paradigm. Tilly, Chris, Rina Agarwala, Pun Ngai, Carlos Salas, and Hina Sheikh. 2013. Informal Worker Organizing as a Strategy for Improving Subcontracted Work in the Textile and Apparel Industries of Brazil, South Africa, India and China. International Labor Affairs Bureau (US Department of Labor) and UCLA Institute for Research on Labor and Employment, November 2013. https://irle.ucla.edu/wp-content/uploads/2017/05/Informalworkerorganizing intextilesandgarments-UCLAReport-9-2013.pdf (accessed March 14, 2021). Tilly, Chris, Georgina Rojas-García, and Nik Theodore. 2019. Intersectional His- tories, Overdetermined Fortunes: Understanding Mexican and US Domestic Worker Movements. Political Power and Social Theory 35: 121–45. Tilly, Chris and Georgina Rojas-García. 2021. Shifting Tides of Informal Worker Resistance in Mexico: A Domestic Work-Construction Contrast. Bulletin of Latin American Research, 40(3). DOI: https://onlinelibrary.wiley.com/ doi/10.1111/blar.13109, 2021 (accessed March 14, 2021). Vosko, Leah F, Ed. 2006. Precarious Employment: Understanding Labour Market Insecurity in Canada. Montreal, QC and Kingston, ON: McGill-Queens University Press. Weil, David. 2008. ‘Mighty Monolith or Fractured Federation? Business Oppo- sition and the Enactment of Workplace Legislation.’ In Annette Bernhardt, Heather Boushey, Laura Dresser, and Chris Tilly (Eds.), The Gloves-Off Econ- omy: Labor Standards at the Bottom of the American Labor Market. Ithaca, NY: Cornell University Press. Weil, David. 2017. The Fissured Workplace: Why Work Became So Bad for So Many and What Can Be Done to Improve It. Cambridge, MA: Harvard University Press. WIEGO. 2020a. About us. Women in Informal Employment: Globalizing and Organizing (WIEGO). December, 2020, https://www.wiego.org/about-us (accessed August 25, 2020). WIEGO. 2020b. Global Movement of Home-Based Workers. Women in Informal Employment: Globalizing and Organizing (WIEGO). December, 2020, https:// www.wiego.org/global-movement-home-based-workers (accessed August 24, 2020). Wright, Erik Olin. Working-Class Power, Capitalist-Class Interests, and Class Compromise. American Journal of Sociology 105(4): 957–1002. 71 Zatz, Noah. ‘Working Beyond the Reach or Grasp of Employment Law.’ In Annette Bernhardt, Heather Boushey, Laura Dresser, and Chris Tilly (Eds.), The Gloves-Off Economy: Labor Standards at the Bottom of the American Labor Market. Ithaca, NY: Cornell University Press. Zhou, Mujun. 2020. NGOs, Ideological Power Networks, and the Rise of Medi- ating Labor Publics in China. Unpublished paper. Department of Sociology, Zhejiang University. Part II Digital Transformations 75 4 The Increasing and Invisible Impact of the Working Consumer on Paid Work George Ritzer and Piergiorgio Degli Esposti 4.1 Introduction The world of work and labour is being transformed in many ways by an array of well-known and documented forces. Much has been written about automation and robotization and how they have eliminated a multitude of tasks and dramatically changed many of the remainder. Here we focus on a largely ignored force that is having much the same kind of effect on the world of work and has served to amplify the effects of these other changes; the increasing importance of one type of prosumer, the ‘working consumer’. While consumers have always worked, a series of relatively recent changes (i.e. new self-service technologies; the explosion of consumption on the Internet) have served to highlight the importance of the working consumer, prompt- ing a recent expression of concern for the ‘overworked consumer’ (Andrews 2019). Increasingly today, the consumer has little choice other than to work in order to consume. As a result, workers (this time the ‘consuming producer’ another type of prosumer) have become less significant in those contexts in which working consumers have been of growing importance. The consuming worker should be understood as a traditionally paid worker, but one who naturally consumes in a consumption-based economy. Many workers have lost their jobs because of the increase of tasks undertaken, both consciously and unconsciously, by working consumers. The basic principle is that the increasing amount of work being carried out by consumers rather than by paid employees is a largely invisible as- pect of the ‘work revolution’. These consumers offer many advantages over workers, one of which is that they often work for little or no pay. There is a growing requirement for working consumers which impacts, both negatively and positively, the paid worker market. Often blind to this impact we tend 76 to think of producers (workers) and consumers in binary terms. In fact, all consumers must inevitably do some work and producers must inevitably consume. If we think in this way, we can begin to see that consumers have done and currently do an increasing amount of work which was at one time and in different contexts carried out by paid workers. Previously, prosumption has been considered as the fusion of produc- tion and consumption (Ritzer and Jurgenson 2010; Ritzer 2014; Ritzer 2015). Here we move away from this generalization to focus on one type of prosumer—the ‘working consumer’ (Dujarier 2016; Rieder and Voss 2010). The working consumer can be defined as an increasingly important sub-type of prosumption, as the consumer takes part in several parts of the production while consuming, often unknowingly, with only achieving little, if any earnings. The first object of this chapter is to explore the implications of the working consumer on paid employment. Afterwards, the issue of a market for working consumers will be considered, including their effect on the paid labour market (or so-called ‘consuming workers’). These distinctions are based on the view that both produce and consume (prosume), although they do so in different ways. The working consumer produces in order to consume whereas the consuming worker consumes in order to produce. It is argued that working consumers cannot be considered to be a part of a labour market, but that their work nevertheless is getting into conflict with paid labour. This has a series of consequences for both types of labour, which will be discussed conclusively. 4.2 Prosumption and Working Consumers Although prosumption (including working consumption) has always existed, in the contemporary world it is now taking on many new forms creating the ‘new prosumer’. The working consumer is located along a continuum between production and consumption. This is true of both bricks-and-mor- tar consumption settings such as McDonald’s and IKEA but also of online digital sites like Amazon, Facebook, and Google. The fusion of production and consumption as well as that of digital and material is augmented in settings involving both the digital and material. Amazon, for example, sup- plements its powerful online presence with bricks-and-mortar settings such as its Whole Foods supermarket chain and Amazon Go convenience stores thereby producing and consuming in the online digital world and the material world at the same time. 77 Although some research on the working consumer and similar ideas has been done, more attention has been devoted to general issues of prosumer- ism. While these concepts are almost unknown in popular literature, many scholars have been using them and other overlapping concepts for decades. In the past, other scholars have dealt with the process without using the term prosumption. In fact, the phenomenon itself is not new and is undoubtedly more primordial than either production or consumption. Hunter-gatherers could be seen as prosumers as they produced their own food and then con- sumed it, maybe even consuming it as was produced. People were prosumers before they were defined as such classing themselves as either producers or consumers. This distinction gained traction with the Industrial Revolution as many people left home or the farm to work in production settings (i.e. workshops, factories). More recently the Consumer Revolution has brought a sense of people as consumers with the development and proliferation of specific sites (i.e. stores, shopping malls) where people go to consume. As a result, both in the past and now, scholars and laypeople have tended to analyse the economy focusing on production or consumption, as well as workers or consumers, a situation which requires correction. Concern with prosumers in general and working consumers, in particular, should help correct that error. While we have always been prosumers or, more specifically, working consumers, the increasing fusion of work (production) and consumption today is much clearer to both casual observers and scholars who have created and expanded upon these concepts. Other authors have dealt with similar or closely related phenomena such as produser (Bruns 2008), ‘co-creation’ (Prahalad and Ramaswamy 2004); marketing’s ‘service-dominant logic’ (Vargo and Lusch 2004), ‘wikinomics’ (partly based on the idea that businesses put consumers to work on the Internet), ‘DIY’ (Fox 2014), ‘craft consumption’ (Campbell 2005), ‘mass collaboration’ (in both production and consumption) (Tapscott and Williams 2008), and ‘consumer co-production’ (Etgar 2008), makers-fixers-sharers and testers (Degli Esposti 2015). Although these ideas and others, such as the consumer as the manager of workers on sites such as Yelp (Andrews 2019), overlap and have their strengths it is the idea of the prosumer that has been most influential in social science and in our work. In her work on surveillance capitalism, Zuboff (2019) makes the point that in their relationship with Internet sites like Google and Facebook peo- ple cannot be seen as customers as there is no economic exchange, price or profit, but neither as workers as they are not paid for what they do. If they are neither consumers nor producers (workers), perhaps it is better to see them for what they are, prosumers. 78 The Digital Revolution has turned the dualistic thinking of the Industrial and Consumer Revolutions on its head. These tended to focus on either the producer (worker) or the consumer. Such binary thinking is impossible to sustain, especially on the Internet, where it becomes virtually impossible to separate the deeply intertwined acts of production and consumption. For example, in writing on someone else’s Facebook page (production), one must have consumed information on that page or elsewhere. The Digital Revolution has sounded the death knell (although it should have been tolled long ago with the rise of postmodern theory) for modern binary thinking in general and especially in regard to the economy. Rather than the binaries of production and consumption, there is increasing interest in prosumption where the two processes are fused in various ways and to varying degrees. Contemporary interest and use of the prosumer concept can be traced to Alvin Toffler’s (1980) thinking on the rise of the prosumer as well his later work with Heidi Toffler (2006) on the coming prosumer explosion. This work was only part of Toffler’s broader thinking on social change, especially the third wave or post-industrial society. This idea received attention and popular interest for a time although it did not attract the interest of scholars and find its way into academic literature. Although Toffler’s work on prosumption was unseen by most scholars the idea and phenomenon was recorded in the study of McDonald’s and its broader influence through the ‘McDonaldization of society’ (Ritzer 1983; Ritzer 2018). Of particular interest is the way in which McDonald’s, as well as its emulators, extenders, and some predecessors such as earlier chains of fast-food restaurants like Dairy Queen and cafeterias like Automat, put its customers to work in its bricks-and-mortar settings. Customers (consumers) in these restaurants are required to, at least in part, ‘produce’ their own meal by doing work, for example carrying food trays to the table, disposing of garbage when the meal is finished all of which was previously done by paid employees. The line between consumer and worker is therefore blurred, or partly, in fast-food restaurants. This is also now the case in many other bricks-and-mortar settings. Once the traditional department store had many paid workers doing a wide range of tasks for consumers (imagine!). Now with fewer employees consumers do much of the work themselves such as locating purchases among a vast array of products, scanning tags to check or find prices, and scanning purchases when they leave via a self-service lane. Supermarkets still have many employees; however, they are often supplemented by self-service checkout lanes where customers scan purchases, weigh their produce, and bag their purchases. Gone are the days when there were employees available to pump gasoline in 79 service stations. Customers now not only pump (produce) their own gasoline but also pay for it by scanning their credit cards. Customers increasingly use self-check-in kiosks at hotels and airports. They are more and more likely to have to find their own cars in car rental lots, wash their cars at automated car washes, and check out their selections in libraries. IKEA’s customers must not only trek through its seemingly endless maze on their own in an effort to find what they are looking for while discovering and selecting other products in the course of their rambles, but must also assemble products such as bookcases purchased in store at home. Perhaps the epitome in the use of the working consumer in the material world so far can be found in Amazon Go’s convenience stores (ten had been opened by early 2019 with as many as 2000 planned). Amazon Go is at the forefront (another being salad chain Sweetgreen, [Zraick 2019]) of efforts by bricks-and-mortar shops and malls to better compete with online sites by further increasing the use of working consumers and limiting the number of paid employees. As a result, customers are forced to perform work tradition- ally done by paid employees. The Amazon Go ‘grab-and-go’ system allows consumers to enter the bricks-and-mortar shop and to quickly and easily make their selections (groceries, ready-to-eat meals, meal kits, etc.). Because of Amazon Go’s extensive use of digital technology there is no need to wait in line to pay for purchases as it offers checkout-free shopping via the Amazon Go app. Uber has done much the same thing as rides are prepaid through an app. Passengers can exit an Uber without the need to pay or tip. Other shops and malls are likely to follow the Amazon Go model by utilising technology that recognises customers and their preferences as soon as they enter leading them to likely sites and products. Amazon will likely integrate its Amazon Go stores, Whole Foods’ super- markets, and bricks-and-mortar book stores into its far more important digital business. It may use its distribution centers for digitally ordered products or as launch pads for its nascent drone-delivery system. Amazon is expanding in a multitude of directions, augmenting its online business in so many different ways that it has created the fear of monopoly similar to that of the 19th-century railroads that led to the development of anti-monopoly laws. We are clearly in the early stages of the emergence of augmented businesses involving ever-tighter integration of the digital and the material and the degree in which they strengthen one another. In addition to the use of drones other advances being considered are shops staffed by robots using facial-recognition software and the use of 3D printers (or additive manufacturing), which allow consumers to materially produce what they will consume. 80 These advanced technologies highlight the major role that they play in enabling working consumers to do things (e.g. manufacturing products with 3D printers) that in the past could only be done by paid employees. While the working consumer remains important to the existence and de- velopment of bricks-and-mortar businesses, as well as to those that integrate bricks-and-clicks, the most important contemporary examples of the increasing centrality of the working consumer are to be found on Internet sites such as Google, Facebook, and Amazon, as well as more specific sites like TurboTax and LegalZoom. It is virtually impossible to interact with human employees on most Internet sites, including those that sell goods and services, because work done by paid employees is comparatively expensive, prone to mistakes and is unreliable. The near-total absence of human employees online is also due to the fact that most online work is performed by advanced technolo- gies. Additionally and central to this argument is that online consumers do a lot more unpaid work which is not required of them in bricks-and-mortar settings. Often they have no choice but to do such work. Amazon working consumers, for example, must do all of the digital work involved in ordering the myriad products available on their website and many others like it. In the case of books, buyers who select on the basis of online reviews written by other working consumers, can also rate and produce reviews of other books themselves. Increasingly, these working consumers may even author the digital books for sale on the site. As a result of all of the work being car- ried out by its working consumers Amazon has less need for paid employees such as clerks and book reviewers although it does employ many thousands of people who, stifled by routines, rules, and metrics, work like robots in its distribution centres as well delivering products to its working consumers (Scheiber 2019). The increasing power of Amazon in many sectors is forcing a large number of bricks-and-mortar shops, most notably bookshops, out of business with consequent job losses and unemployment. 4.3 Digitalization of Labour Digital processes and technologies are thought of as intrinsic to contemporary society offering many benefits as they innovate or re-envision daily processes. The digital era adds value to how people navigate life from shopping for es- sentials to booking a vacation. The incorporation of digital technologies has enabled companies to develop new services as is done by many tech start-ups, or has allowed conventional bricks-and-mortar businesses to remain compet- itive. Companies of all types have become integrated with digital technology 81 in varying degrees. While some businesses remain as either physical (brick- and-mortar) or digital (social media) the vast majority evolve into hybrids. For example, many retail companies allow consumers to visit their stores in person and/or engage with an online catalogue and purchase system. The nature of employment is also evolving from labour once rooted in the physical or material world, through the development of Internet communi- cation technologies (Kässi and Lehdonvirta 2018). Remote working, fixed- hour work arrangements and project-based contracts are now commonplace. Platform-based companies such as the Big Five (Google, Apple, Facebook, Amazon, Microsoft) are not merely digital technologies that facilitate various activities for users, but they are technological, economic, and social-cultural configurations (Van Dijck, de Waal and Poell 2018). These platforms collect information about user behaviour and preference generating an enormous amount of data, or big data, which is then processed by algorithms which in turn aid the development and creation of upgraded products and ser- vices (Dawar 2016). Application of this data is seen as a marketing tool for customer relationship management providing platforms with a competitive advantage (Dawar 2016). Utilization of big data and algorithms also drives the platform’s profitability and many decision-making elements. Algorithms are not only used to manage the work that is performed but are used to control almost all aspects of the job, including worker performance and behaviour (Marrone and Finotto 2019). Digital platforms are also developing into sites for work which is entirely web-based and/or physical labour allocated through the platform (Berg et al. 2018; Gajewski 2018). Digital labour has two distinct types, crowd work and online freelancing. Crowd work tasks are subdivided into smaller work units such as data entry or survey completion. Contest-based work refers to such things as designing a logo that is voted on by an entity that requests the task (Schmidt 2017). Through online freelancing individuals are selected to complete more substantial tasks such as editing, translation, or graphic design (Schmidt 2017). Although digital labour can be divided into these two types in reality the division is blurred with digital labour being seen as a continuum of online labour from microwork to online freelancing (Schmidt 2017). Some digital, if not all, platforms have ‘attempted to evade existing labour regulations’ as they rely upon being considered as a new or untraditional form of labour (Berg et al. 2018, 6; Jourdain 2018, 6). On sites that require users to supply or generate content, particularly social media sites, content moderators are needed to monitor uploaded content and to remove unacceptable material. Although algorithms are also used to assess questionable content, the judgment of human workers is still essential for quality control (Berg et al. 2018). 82 The three business forms, brick-and-mortar, brick-and-clicks, and digital platforms, each have their own characteristics relating to goods and services offered as well as their labour needs. A business may enter the market primarily in a certain form but as it continues to adapt to remain competitive it may mutate into a different or hybrid form fluctuating between the physical and digital. Changes wrought by the digital era have impacted, not only business development, but entire industries and markets as well as shaping consumer participation and engagement. The roles of consumer and worker have also become muddled as they flow between varying degrees of prosumption as required by the company’s structure. The nature of paid work itself is changing dramatically with an increasing shift away from solid jobs in material settings to more liquid, virtual work woven between physical and digital spaces. Workers are increasingly moving out of traditional full-time jobs in large organizations, either by choice or force, into the on-demand gig economy in which people regularly move from one short-term position to another. With the rise of platforms workers are increasingly employed outside standard regulations, being characterized by low wages, exclusion from welfare protection and also from the possibility to get access to traditional union rights (Marrone and Finotto 2019). Moreover, platform’s exploitation seems to not be limited to the same workers, but it also affects the same context in which they operate. Platforms, in fact, are also able to influence social life in the cities as they are increasingly becoming crucial infrastructure upon which urban economy relies (Marrone and Peterlongo 2020). Many factors allowing such changes are only possible through varying layers of exploitation, intricately employed to increase the profits of companies. The digital era is often thought of as increasing employment opportunities through, for example, the gig economy, however, this innovation has increased worker exploitation through unstable jobs which are less financially rewarding and are often accompanied by poor working conditions. Those impacted most by this exploitation are those whose involvement is fundamentally required by the profit-making company: the worker and the consumer. Users, or consumers, are required to interact with digital platforms to request services such as booking a room on Airbnb or hiring someone to assemble furniture through TaskRabbit, thus performing a type of unpaid work. The use of hyper-logical algorithms as surrogate su- pervisors of paid workers employed by digital labour platforms often results in unfair worker assessment and treatment (Berg et al. 2018). 83 4.4 Digital Platforms and the Gig Economy While prosumption and the working consumer have helped profit-making companies reduce labour costs digitization has facilitated this process. The digital era has enabled entrepreneurial profit-making companies to recognise and create niche markets that can be monetized and made profitable through digital technologies. Although they promote services as tools to benefit the consumer they emanate from within the economic system itself and their priority is to make a profit. These companies have identified consumption practices previously performed by the working consumer for free which do not require specialized skills to complete. As individuals raise their standards as to what constitutes decent work many tasks outside of their paid working life are outsourced to others through the gig economy. A contradictory flow is occurring in the labour market. Structural pro- cesses are being introduced that encourage or require unspecialized working consumers to perform many tasks once performed by paid consuming workers causing many low-skilled workers to be laid off. These laid-off workers then become job seekers and gig-based companies are in need of cheap labour. With the rational goal of efficiency in mind these new businesses seek to employ unskilled workers willing to perform low-level tasks for low pay and poor working conditions. Through this shift contracted gig workers become a specific subset of the P-A-P (Prosumer as Producer) continuum (Ritzer 2017), sharing many attributes with the working consumer particularly the requirement that they use their own assets to perform their jobs. As companies grow accustomed to cost savings and benefits provided by working consumers their ethos of the treatment of the consumer is applied to paid workers. The aim of extracting value is especially evident in the gig economy as capitalist companies exploit workers through poor working conditions and minimum wages in order to maximize their labour savings. Gig economy working conditions, like that of low-level consuming workers, are generally considered indecent and dehumanizing. There are relatively few benefits with irregular and/or non-guaranteed hours, little or no training, and limited possibility of career advancement (Muntaner 2018). The nature of the work is also alienating, similar to that of the online working consumer as many activities are performed in solitude, with little or no contact with those requesting the service and limited contact with co-workers. Underutilization of the gig worker perpetuates this feeling of loneliness (Graham et al. 2017). As the work is intermittent the time between tasks is often accompanied by long periods of idleness, which adds to the sensation of restlessness and isolation. Importantly, gig work does not provide employees with work that 84 could be considered to be meaningful. Similar to Amazon employees, gig workers are required to juggle multiple bosses: their immediate supervisor, the algorithm that dictates the amount of work and how the process must be performed, performance metrics and consumer ratings (Bajwa et al. 2018). The relationship between the digital and physical increases the pressure as well as the level of possible exploitation on the worker. While jobs in the gig economy are recognizably demeaning being unem- ployed is much worse, therefore, the unemployed or under-employed settle for these poor conditions in order to receive at least some pay. Since many such jobs are contract positions pay varies greatly depending on demand. Many potential gig workers are given false promises that the work will be lucrative so they take the risk. Therefore, supply meets demand and low-level workers begin new, demeaning jobs with a gig-based company seeking to minimize its labour costs. The increasing prevalence of work in the gig economy also makes it dif- ficult for gig seekers to find and keep employment. Unless job seekers are highly educated, skilled, or have unique abilities, they have little power in the labour market as they are easily replaceable and unlikely to have anyone representing their interests. Despite the associated problems or dissatisfaction an increasing number of people will be employed in the gig sector in the future. It is true that some gig economy jobs offer empowerment, autonomy, a wide variety of tasks, networking, and working schedule flexibility which renders the job meaningful and decent. Certain working environments of gig work also offer an alternative to the restrictiveness of conventional corporate workplaces. Some types of gigging enable greater personalization of both the workplace and the work. 4.5 Of Unemployment and Employment Because of the increasing importance of working consumers people are losing their jobs, being forced into part-time work, are under-employed, or not getting paid jobs at all. Although there are no hard data supporting this yet, it is a logical argument. Several studies have shown how automatization and robotization are affecting the future labour market, putting low-skilled employees out of work (e.g. Frey and Osborne 2013; Manyika et al. 2017). It is clear that working consumers are doing what was once traditionally paid work. It can, therefore, be argued, that working consumers overtaking prior tasks of paid workers, will have similar consequences to those of automatiza- tion. Both developments will simultaneously create other jobs, though often 85 demanding higher levels of skill, which will be discussed later in relation to working consumers. Consumers are working, often seemingly happily, on an unpaid basis or for poor reward. But the advantages of working consumers over paid workers do not stop at being unpaid or poorly paid. These unpaid or poorly paid working consumers offer the profit-making organization many advantages apart from savings in labour costs. Savings are also made in marketing, advertising, and in salespeople used to induce traditional customers. While profit-making organizations still have many short and long-term obligations to paid workers, there are few, if any, responsibilities to working consumers. In addition to paying a wage to paid workers employers may also be responsible for various costly benefit programs such as health insurance, retirement programs, and paid vacations. There are no such responsibilities to working consumers. In addition, paid workers need to be provided with the necessary and often costly ‘means of production’ such as places to work (offices, factories), tools, and machines (assembly lines, computers). By contrast, some working consumers pay for the purchase and upkeep of their own means of production such as home offices, associated utilities, computers, and transportation, if they drive for a ride-sharing company for example. Working consumers also cost less to serve. Fewer paid personnel are needed in shopping sites such as department stores because prosumers now do much of the work themselves. There are even greater savings in terms of the increasingly important consump- tion on Internet sites like Amazon and eBay, and travel sites such as Trivago, KAYAK, and Expedia where there are almost no paid employees, with the unpaid working prosumers doing virtually all of the work. Other savings are derived from the fact that products are either stored by working consumers as is the case with much for sale on eBay and used books on Amazon, or are sold on more of a just-in-time than a just-in-case basis. Amazon does not warehouse the vast majority of the books and other products it offers for sale but rather obtains them from third-party sellers, often themselves working consumers, as they are ordered. These advantages and savings are an irresistible attraction to profit-making organizations which covet both fewer responsibilities and most importantly from the point of view of profits, a great reduction in costs. The future of work will involve more and more activities that are not work. The future will bring with it less paid work and more unpaid activities by working consumers. Obviously, many people will continue to be paid to work and new paid work is being created by the ongoing revolution in the working world. 86 While we have discussed its role in job losses, working consumption also leads to job creation. One of the best examples involves bloggers who turn their activities into paid work by, for example, finding advertisers for their blogs or by using their success as bloggers as a springboard into becoming reporters, book authors, etc. More importantly, working consumption, and prosumption in general, relies on and has led to the creation of millions of new jobs for paid employees. For example, because of the billions of dollars spent by its working consumers Amazon employs about six hundred million paid employees. Then there the innumerable number of workers in companies involved in producing the systems, Smartphones, ATMs, self-checkout technology, websites, and so on that make working consumption on the Internet the norm. It is possible that more jobs are lost as a result of working consumption than are created by it but more importantly is the fact that those who obtain these new paid jobs are unlikely to be the same people who lose their positions as a result of the working consumer. For example, an unskilled supermarket worker or bank teller is not likely to find their way into the high-tech industries that owe their existence, at least in part, to the increasing centrality of working consumption. Those industries often require a more advanced or different skill set. 4.6 The Market for Working Consumers As the former part of the chapter has indicated, it is crucial to conceptual- ize the working conditions of working consumers, due to their increasing importance in contemporary capitalism, producing a large amount of value without realizing it. Given this background, we discuss whether there exists a market for working consumers and the nature of that market (is it a labour market?). Furthermore, we investigate how it relates to the well-established market for paid employees (consuming workers). Beyond the fact that we do not think of consumers as working, we do not imagine a market for consumers per se because they do not meet the criteria for a market. They are not seen as offering any goods or services and they are not paid, at least in a traditional sense, for what they do. Although we may not think of consumers in general as a market, it however makes sense to think of working consumers as being involved in a market as in many cases they are doing work that could be, and often was and is, performed by paid employees. In many cases working consumers are replacing paid workers and therefore should be seen in the same light, at least from a market point of 87 view. Some working consumers can also be seen as being ‘paid’, or econom- ically rewarded, for their work. This is not the case in a bricks-and-mortar setting such as a supermarket where prices are the same whether one uses a traditional checkout lane or ‘works’ on a self-checkout line. It is nevertheless arguable that those who order books online from Amazon for example are ‘paid’ with lower prices. The same is true for the dwindling number of service stations that offer full-service and self-service lanes where those who pump their own gasoline get lower prices than those who do not. The market for working consumers is very different than the general conventional market and also from the market for paid employees. These markets generally consist of two distinct roles, sellers and buyers (Granovetter 1995; Zelizer 1983; Ahrne et al. 2014). Those involved in markets under- stand their distinct roles and the expectations associated with them. They also understand the expectations associated with their relationship to one another. For example, in a labour market workers understand that they are sellers and their employers are buyers and they understand the expectations associated with those roles and relationships. Although binary thinking is involved in both the general conventional market and, more specifically, the labour market, binary thinking is clearly rejected in work on the prosumer, including the working consumer and the consuming worker. With the concept of seller and buyer the working consumer and the consuming worker is almost always both. While they are buying goods and services, working consumers are, in a sense, selling their services to the supermarket or Amazon, albeit for the minimal return of a price lower than would have been the case if they did not ‘work’. These entities in turn sell goods and services and are buyers of the services offered by working consumers which they ‘pay’ for in the form of lower prices. The concepts of buyer and seller are simplistic binary perspectives which erroneously imply that the parties involved fully understand their roles as buyers and sellers. Working consumers do not fully understand that they are working or at least doing tasks similar to those performed by paid workers. They do not regard themselves as sellers nor do those who use their services actually buy them in the conventional sense. Buyers do not fully understand that they are buying the services of working consumers. If neither buyer nor seller fully understands what they are doing, can this therefore be regarded as a conventional market? Within the dualistic perspective of seller-buyer, there is the concept of an economic exchange between them. In many cases involving the working consumer, where no such exchange occurs, this concept does not apply. By definition, working consumers work as they consume often doing so with no 88 economic advantage. Those who use the self-checkouts at the supermarket are an obvious example. Working consumers may read (consume) and write (produce) on another’s Facebook page whilst receiving no economic result for themselves although Facebook increases its economic value each time this occurs. Facebook, Twitter, and others do not pay for the work done by prosumers; however, it uses the income from the burgeoning revenue derived mainly from the tasks performed by working consumers to increase its offers. Or put in another way, by posting and commenting, users produce most of the content on social platforms. This is what keep other users active and leaving data footprints, which are profitable for platform providers. This revenue is not shared with the working consumers who are the main source of the big data that is the ultimate source of its the success of man social platforms. While conventional workers in the kinds of businesses discussed here are relatively powerless (e.g. Amazon warehouse workers), working consumers are much more powerless as they are not fully aware of what they are doing. Working consumers also lack the resources of paid workers and those who use their services. There are laws, contracts, and Unions that strengthen the position of paid workers; however, these do not exist for working consum- ers. As they do not regard themselves as working, they do not demand such things. Working consumers are invisible to the Government, Management, and Labour Unions as they do not regard them as being engaged in labour. This means that these agencies are not available to give them the power ac- corded to paid workers. To summarise, the market for working consumers are not similar to a conventional labour market, because of the fluid roles of buyers and sellers, the lack of economic exchange and invisibility of their working rights. Markets are usually seen as existing outside of organizations, but organi- zations can be seen as existing within markets and markets as existing within organizations although these binary perspectives are problematic. Working consumers are neither clearly in a market outside or inside an organization. For example, is there a market outside of Amazon for its working consumers? In theory, they can take their working consumption elsewhere, to a compet- ing website or to a bricks-and-mortar superstore. However, these businesses struggle competing with the increasingly hegemonic Amazon, including its bricks-and-mortar entities. This serves to make Amazon a market in itself. In general, despite growing augmentation, the markets for working consumers remain separate from one another, especially in regard to digital and bricks- and-mortar markets. The conditions of working consumers, therefore, become less flexible, as they cannot choose the most favourable site for prosumption, but increasingly have to accept rules of hegemonic organizations like Amazon. 89 If working consumers are not clearly in any market, are they part of organi- zations such as Amazon, Facebook, and Twitter? Although increasing integral and crucial to these organizations as both consumers and workers they are decidedly not part of them. They are not recognized as part or members of any organization, are not paid (at least directly) by them, have no workspace, and receive no benefits. Working consumers give an enormous gift to these businesses in that they consume what the business has to offer and do most of the work required to obtain the offering free of charge. Working consumers do all this and more while the entities that they work for owe them little and provide them with little or nothing. This constitutes an invisible exploitation of workers that seems more extensive than the one of industrial capitalism, since working consumers are left unaware of their labour and the connected rights they are entitled to. Working consumers cannot be said to belong to either markets or organi- zations, but they are related to both. Furthermore, as working consumers do not view themselves as such, they do not see themselves as part of a market or as part of an organization such as Amazon. Organizations prefer this because it absolves them of any responsibility or obligation towards these consumers. Markets and organizations have hierarchies which often overlap; the po- sition of the working consumer in these hierarchies, like everything about them, is unclear and ambiguous. While working consumers are critical to the success of organizations such as Amazon, Amazon Go, Facebook, and Google, they do not occupy a clear position in their hierarchies. They are seen, and see themselves, as outsiders. They play a key role in making those hierarchies possible and economically successful but are not seen, and do not see themselves, as part of these hierarchies. They are ‘outsiders within’ (Collins 1986, 26). As outsiders, they play little or no role in the creation and administration of the markets and organizations in which they operate. These are organized by other major players in the markets and those at or near the top in organi- zations. While they are essential to many markets and organizations they must function in accordance with ground rules created and managed by others. It would seem that working consumers fit better with ‘switch-role’ rather than fixed-role markets although they do not so much switch roles, in for example from buyer to seller and back again, but rather hover between both. The switch-role idea, however, is undermined by its binary thinking. Worse is the notion of ‘fixed-role markets’ since working consumers never occupy a fixed role as they are never either buyers or sellers but to varying degrees always a combination of the two. 90 Markets typically operate within formal and informal rules; however, there seem to be few if any of any of these which involve the relationship between working consumers and the settings in which they work. Again, this is be- cause consumers do not see themselves as working and the settings in which they work do not see them as workers. Given these vagaries little in the way of rules concerning working consumers has emerged. Rules that exist have been created one-sidedly by those who run the settings and are designed to work to their advantage. Working consumers have created very few rules as they are largely unaware of what it is they are doing. Their position on the periphery of everything, especially markets and organizations, gives them little power to do anything, especially create the rules by which they operate. Those who occupy the role of working consumer are sometimes seen as being without qualities, with many different, loosely defined qualities (Musil 1930) or lacking in extrinsic relationships (Helmer 1970). Without intrinsic relationships people also lack ‘qualities’, depth, a centre. In terms of this discussion, as they do not regard themselves as working consumers and they have no extrinsic relationships, they are unable to see themselves as having qualities, at least in their role as working consumers. Instead, they rely on others, especially markets and organizations, to define them and their qualities, or lack thereof. As they are largely invisible to both markets and organizations their definitions are vague and ambiguous. While buyers and sellers can be seen as having agency (being conscious and creative about their actions), can the same be said for working consumers? As they frequently combine elements of buyer and seller, there must be at least some agency involved in what they do. However, because working consumers do not regard themselves as buyers or sellers, let alone working consumers, can they be seen as being agents if they are not conscious of their role? As working consumers by definition work, can they be seen as part of a labour market? A labour market is defined as an arena in which people offer their labour-power in exchange for wages, status, and other job-related rewards (Kalleberg and Sorenson 1979). Working consumers, however, do not receive wages or other economic or job-related rewards. This is partly due to the fact that they are not regarded as holding jobs and therefore not part of the labour market. Although they labour, working consumers are largely seen as consumers and not workers. Conscious of this Management relies increasingly on working consumers rather than paid employees as they are not seen as workers, receive no wages and, more broadly, because of the associated cost-savings. In fact, working consumers pay for the privilege of doing their work as they buy and use their own equipment as well as pay for the products and services available at the businesses they visit. 91 The labour market is particularly segmented and stratified. Some workers exist in a favourable labour market whereas many others are found in a less favourable or unfavourable segment, both of which have many gradations. The unemployed and underemployed are in a whole other segment and are even more disadvantaged. The growing number of gig workers (e.g. Uber drivers) may be thought of as belonging to yet another segment. Working consumers have never been thought of as labour, let alone in a labour segment, and would therefore need to be defined in yet another segment or category, one that has never been previously considered. A primary favourable labour segment offers high wages, job stability, good working conditions and solid opportunities for advancement, whereas a secondary unfavourable segment the antitheses of these and other vari- ables. There is great concern for the plight of those who are employed in a secondary segment. Working consumers would seem to occupy yet another, as yet uncategorized, segment which also has many even more unfavourable characteristics. In fact, it may even be less desirable than that of those who are poorly paid, in unstable positions, with poor working conditions, and little opportunity for advancement. In general working consumers are paid little or no wages, and there is little stability as they move from one consumption setting, either bricks-and-mortar or digital, to another. Another negative of their working conditions is that they often do low-skilled work, sometimes on a kind of assembly-line or at check-in/out counters or in the work they do online. Even though they work online, they are set apart from the online labour market, which includes those who labour for pennies on such systems as Amazon’s Mechanical Turk. There is also no opportunity for advancement for working consumers either on or off-line. The whole idea of advancement for a working consumer is ludicrous—being ‘promoted’ to a first-rate user of Amazon.com or a champion user of the self-service checkout lane? In bricks-and-mortar settings, paid employees and working consumers constitute a kind of internal labour market in which they are sometimes in competition with each other. Because they are not paid, preference goes to working consumers overpaid employees often costing them their jobs, pushing them into part-time jobs and thereby reducing their income. This cutting-edge approach may lead many profit-making organizations to shift as much work as possible from paid workers to working consumers as there are great economic savings to be had in relying on unpaid working consumers rather than paid workers. Although working consumers may obtain status in a variety of ways by serving as preferred reviewers for Amazon or acquiring a large number of likes on Facebook, in the main they are performing thankless tasks which offer 92 little or nothing in terms of status. For example, they may get approbation from fellow consumers on full-service lines in supermarkets for working for nothing while they themselves are having these tasks done for them by paid cashiers and baggers. Marx and Blauner outlined the concept of alienation in terms of power- lessness, meaninglessness, isolation, and self-estrangement. Although work- ing consumers are not as powerless as lower-tier paid employees they have little power in bricks-and-mortar settings for example, where they have no choice but to use self-service systems and are propelled along them in a kind of assembly-line fashion. Online the situation is even worse where working consumers operate in a mindless fashion in ways dictated by the system. It would be difficult to think of tasks like pumping one’s own gas as meaning- ful. Working consumers also operate in isolation from other consumers who hold paid jobs and they may be seen as self-estranged as they mindlessly do a variety of tedious tasks both off and online. Although working consumers can be thought of as alienated they are not as alienated as the now disappearing assembly-line worker. In fact, they do not regard themselves as alienated because they do not see themselves as workers. Their problems are largely invisible, built into the structures within which they operate; they are not social-psychological. Some working consumers do obtain a type of status, however, at the moment it is impossible to think of them as having jobs and being a part of the labour market although, as paid employment declines and working con- sumption increases, a future labour market may include working consumers. It is possible that in the future we will need to pay people to be working consumers giving them a variant of a guaranteed annual wage not only for the work they do but also in order to keep the economy operating at a high level. This may make the idea of a guaranteed annual wage more palatable than now because working consumers can be seen, in part, as working. Currently, working consumers are not seen as part of the internal labour market of organizations such as Amazon as they are not internal to it. This results in alienation and the lack of favourable working conditions. Never- theless, this development does not seem to get reversed, since organizations profit from the missing recognition of consuming work. However, they do labour that is very similar to that carried out by paid workers; therefore, this may need to be reconsidered in the future. Working consumption should be seen as labour and the working consumer regarded, at least in part, as internal to such organizations as they do an ever-increasing amount of the work that takes place in them requiring payment (at least in some sense) of a wage. 93 4.7 The Relationship between the Markets for Working Consumers and Paid Workers 4.7.1 Two Separate Peacefully Co-Existing Markets While it is certainly the case that not all areas of the economy involve paid workers, working consumers, those shopping in bricks-and-mortar and on- line settings, are involved. It is not hard to think of paid workers as existing in a market, especially a labour market; however, it is unusual to think of consumers in that way. Once we combine ‘working’ with ‘consumers’ it is much easier to think of them as existing in a market. The markets for paid workers and largely unpaid working consumers in these areas of the economy, although interlinked in various ways, can be seen as being separate markets existing alongside one another. While paid workers are typically seen as being involved in a labour market, consumers, whether or not they are regarded as working, are not. In fact, (working) consumers are seen as being served by those who exist in that labour market. Even though working consumers are not seen as functioning in a labour market, although this may change in the future, they may be viewed by others (e.g. sellers) as a market for their products and services. Working consumers can be seen as being in a separate labour market that peacefully co-exists with that of paid workers. Although working consumers are not necessary conscious about their role in the market, paid workers know very well about the existence of working consumers and encourage them, on the one hand, because it means a significant reduction in work and on the other because it means a gain for the company and therefore a ‘guarantee’ of the conservation of their paid work itself. Certainly, up to now, this has been the case. While all consumers must by necessity work it is only in recent years that structural changes have put them in conflict with one another. In the past, paid workers and working consumers were in a generally harmonious relationship with one another. Paid workers provided goods and services and working consumers took and paid for them. While these consumers carried out actions that could be considered as work, such as studying advertisements and price lists or transporting their purchases home, it was generally not work that overlapped with or threatened the work being done by paid employees. Working consumers still engage in such actions today as they study ad- vertisements and price lists although they are now more likely to be found online. In general working consumers continue to co-exist peacefully with paid workers; however, contemporary developments have brought the two 94 into conflict with one another. For example, in the past consumers had to trek to the restaurant to pick up their take-out orders and the work involved did not pose a threat to paid workers but with the advent of online sites as Grubhub and Uber Eats, working consumers who pick up their own food are doing the job of the delivery person free of charge meaning that they are in direct conflict with paid workers. 4.7.2 Two Markets in Conflict with One Another The consumers of interest here are working whether or not they are being paid or otherwise economically rewarded for their work. They exist in a market and that market exists alongside that for paid workers. In some ways, these markets complement each other but of greater interest and importance is the degree to which they compete or conflict. The two-sided market no- tion became very popular in literature with the development of credit card payments, the videogame industry and especially the rise in social network platforms. These two types of agents (paid and unpaid workers) interact on pre-existing free platforms. Agents value positively the presence of those of the opposite type but may value negatively the presence of agents of their own type. Some new platforms introduce fees and subsidies so as to divert agents from pre-existing platforms and make a profit (Belleflamme and Tulemonde 2009). Broadly speaking a two-sided market is one in which two sets of agents interact through an intermediary or platform and the decisions of each set of agents affect the outcomes of the other. Two-sided markets often generate a situation where members of one side use a single platform and the other side uses multiple platforms (Armstrong 2006). Most products and services that have redefined the global business landscape in recent times tie together two distinct groups of users in a network. These users serve what can be called a two-sided market or two-sided network (Eisenmann et al. 2006). Platforms consist of two-sided networks which bring together the two groups of users seeking products and services. These platforms have infrastructures and rules that facilitate the transactions of both groups where costs and revenues are generated by both sides (Eisenmann 2007). The market for working consumers constitutes an alternative to and can be in conflict with the market for paid employees, or at least to the positions of some of those who operate within it. This view can be questioned, however, because it operates with the kind of binary perspective that has already been critiqued on several occasions here. 95 4.7.3 Market Fusion As previously mentioned, the paid worker and consuming worker markets operate with a strong binary perspective; however, if we look at this issue from the point of view of a more integrative approach which sees the fusion of producer and consumer in the prosumer, we can begin to regard paid workers and working consumers as existing in a single market. This has powerful implications for both and their relationship to one another. In fact, although unknowingly, working consumers and paid workers exist in and will increasingly constitute a single market in which they will become ever more interchangeable. The rise of platform capitalism is enabling a new set of alternative forms of enterprise. In the absence of guiding social principles to balance progress and profit, these new enterprises are unlikely to fill the gap in social needs left by the collapse of corporations. So far, creative destruction has been big on destruction but not so great on creation, at least as far as jobs and social welfare are concerned (Davis and Shibulal 2018). We now live in the midst of a new and particularly powerful phase of creative destruction, one that involves rapid evolution of digital technology, swift proliferation of new digitally-based organizations, and the destruction of various business sectors and their enterprises rooted in the material world. New technologies are at the base of many economic changes brought into existence and deployed by capitalist enterprises and entrepreneurs. As a result of these changes, new economic forms (i.e., home or ride-sharing) are pitted against older forms (hotel chains, taxicabs) with the increasing possibility that the new forms will win out, changing or obliterating the older alternatives (Ritzer and De- gli Esposti 2020). Moreover, although creative destruction should involve creative creation, from the moment in which transformations in the market take place at an uncontrolled speed, creative creation lags behind, leaving as a consequence a lake of workers and consumers who must rediscover their role within of the market. Creative destruction is not a minor, quantitative or incremental change but involves major, qualitative, and revolutionary changes. Market fusion may involve the creation of non-proprietary, open-source software platforms that can be used to support local cooperatives in which the role of the working consumer is essential. These operating systems are relatively cheap and efficient and can be applied quite broadly benefiting those that need to accomplish tasks cost-effectively and on a grand scale. 96 4.8 Conclusion The importance of theorizing the working consumer has been underlined, as it is beginning to form a significant role in contemporary production. The concept is rather unknown in academic literature and is dominated by a questionable binary thinking of the prosumer as a generalized term. It seems particularly important to explore the invisible exploitation of working consumers, since companies do not owe them any long-term obligations, such as providing health insurance, decent wage, or retirement plans. This is naturally beneficial to companies’ reduction in costs but is problematic for consumers who remain unaware of their work and its connected rights. Moreover, we investigated whether there exists a market for working con- sumers, since they are replacing paid workers and therefore could be seen in the same light from a market point of view. Working consumers cannot be considered to be a part of a traditional labour market, with its fluid roles of buyers and sellers, the lack of economic exchange and invisibility of work- ing rights. They are neither seen as internal to organizations, where they are excluded from established hierarchies and not given any status or rights. As there are no prospects for advancement or meaningfulness, working consum- ers will feel alienated from their tasks. Nevertheless, there is a purpose of relating a market of working consumers to one of paid workers, as it enables an understanding of future labour. The two markets can be perceived have separate markets, that has been coexisting rather peacefully, but now are in- creasingly conflicting, as they contain similar tasks. But through an integrative perspective, we can understand both actors in a single market, where they will become even more interchangeable than they are now. There are major changes unfolding in the labour market, where the con- sumer is working even more than was true historically. Maybe their tasks will be paid in some sense in the future. But as this chapter has suggested, it will become crucial to fill the social gaps created by the development, that is mainly pushed forward by reduction in labour costs rather than fairer conditions. 4.9 References Ahrne, Göran, Patrik Aspers, and Nils Brunsson. 2014. The Organization of Mar- kets. Organization Studies, 36(1): 7‒27. doi.org/10.1177/0170840614544557. Andrews, Christopher K. 2019. The Overworked Consumer: Self-Checkout, Su- permarkets, and the Do-It-Yourself Economy. Lanham, MD: Lexington Books. Armstrong, Mark. 2006. Competition in Two-Sided Markets. The RAND Journal of Economics, 37(3): 668-691. 97 Bajwa, Uttam, Denise Gastaldo, Erica Di Ruggiero, and Lilian Knorr. 2018. The Health of Workers in The Global Gig Economy. Globalization and Health, 14(1): 1‒4. Belleflamme, Paul and Eric Toulemonde. 2009. Negative Intra‐Group External- ities in Two‐Sided Markets. International Economic Review, 50(1): 245‒272. Berg, Janine, Marianne Furrer, Ellie Harmon, Uma Rani, and M. Six Silberman. 2018. Digital Labour Platforms and the Future of Work: Towards Decent Work in the Online World. Geneva: International Labour Office. Bruns, Axel. 2008. Blogs, Wikipedia, Second Life, and Beyond: From Production to Produsage. New York: Peter Lang. Campbell, Colin. 2005. The Craft Consumer: Culture, Craft and Consumption in a Postmodern Society. Journal of Consumer Culture, 5: 23‒42. Collins, Patricia Hill. 1986. Learning from the Outsider Within: The Sociological Significance of Black Feminist Thought. Social Problems, 33: S14‒S42. Davis, Gerald F. and Sarojini D. Shibulal. 2018. ‘Taming Platform Capitalism to Meet Human Needs.’ In Subramanian Rangan (Ed.), Capitalism Beyond Mutuality? Perspectives Integrating Philosophy and Social Science. Oxford: Oxford University Press, 207‒222. Dawar, Niraj. 2016. Use Big Data to Create Value for Customers, Not Just Target Them. Harvard Business Review. 16 August, 2016, https://hbr.org/2016/08/ use-big-data-to-create-value-for-customers-not-just-target-them (accessed October 22, 2020). Degli Esposti, Piergiorgio. 2015. Essere prosumer nella società digitale: Produzione e consumo tra atomi e bit. Milano: Franco Angeli. Dujarier, Marie-Anne. 2016. Three Sociological Types of Consumer Work. Journal of Consumer Culture, 16: 555‒571. Eisenmann, Thomas R. 2007. Managing Proprietary and Shared Platforms: A Life-Cycle View. Cambridge, Mass: Harvard Business School. Eisenmann, Thomas R., Geoffrey G. Parker, and Marshall W. Van Alstyne. 2006. Strategies for Two-Sided Markets. Harvard Business Review. 1 October, 2006, https://hbr.org/2006/10/strategies-for-two-sided-markets (accessed October 15, 2020). Etgar, Michael. 2008. A Descriptive Model of the Consumer Co-Production Process. Journal of the Academy of Marketing Science, 36(1): 97‒108. Frey, Carl B. and Michael Osborne. 2013. The Future of Employment: How Susceptible are Jobs to Computerisation. Technological Forecasting and Social Change, 114: 1‒77. Gajewski, Eltje. 2018. Ambivalente Flexibilität: Crowdworker*innen zwischen selbstbestimmter Arbeit und kapitalistischer (Selbst-)Verwertung auf digitalen Plattformen. Duisburger Beiträge zur soziologischen Forschung. 3. November 2018, https://www.uni-due.de/soziologie/duisburger_beitraege_dbsf-2018-03. php (accessed October 19, 2020). 98 Graham, Mark, Isis Hjorth, and Vili Lehdonvirta. 2017. Digital Labour and Development: Impacts of Global Digital Labour Platforms and the Gig Economy on Worker Livelihoods. Transfer: European Review of Labour and Research, 23(2): 135‒162. Fox, Stephen. 2014. Third Wave Do-It-Yourself. Technology in Society, 39: 18‒30. Granovetter, Mark. 1995. Getting a Job. Chicago: The University of Chicago Press. Helmer, John. 1970. The Face of the Man Without Qualities. Social Research, 37: 547‒579. Jourdain, Anne. 2018. Analysing the Symbolic Economy with Pierre Bourdieu: The World of Crafts, Forum for Social Economics, 47(3‒4): 342‒361. Kalleberg, Arne L. and Aage B. Sorensen. 1979. The Sociology of Labor Mar- kets. Annual Review of Sociology, 5(1): 351‒379. Kässi, Otto and Vila Lehdonvirta. 2018. Online Labour Index: Measuring the Online Gig Economy for Policy and Research. Technological Forecasting and Social Change, 137: 241‒248. Manyika, James, Michael Chui, Mehdi Miremadi, Jacques Bughin, Katy George, Paul Willmott, and Martin Dewhurst. 2017. A Future that Works: AI, Auto- mation, Employment, and Productivity. McKinsey Global Institute Research. January, 2017, https://www.mckinsey.com/~/media/mckinsey/featured%20 insights/Digital%20Disruption/Harnessing%20automation%20for%20 a%20future%20that%20works/MGI-A-future-that-works-Full-report.ashx (accessed October 19, 2020). Marrone, Marco and Finotto, Vladi. 2019. Challenging Goliath. Informal Union- ism and Digital Platforms in the Food Delivery Sector. The Case of Riders Union Bologna. Partecipazione e Conflitto, 12(3): 691‒716. Marrone, Marco and Peterlongo, Gianmarco. 2020. Where Platforms meet Infrastructures: Digital Platforms, Urban Resistance and the Ambivalence of the City in the Italian Case of Bologna. Work Organisation, Labour & Glo- balisation, 14(1): 119‒135. Muntaner, Carles. 2018. Digital Platforms, Gig Economy, Precarious Employ- ment, and the Invisible Hand of Social Class. International Journal of Health Services, 48(4): 597‒600. Musil, Robert. 1930. Der Mann ohne Eigenschaften. Berlin: Rowohlt. Prahalad, Coimbatore K. and Venkat Ramaswamy. 2004. Co‐creation Experiences: The Next Practice in Value Creation. Journal of Interactive Marketing, 18: 5‒14. Rieder, Kirsten and Gerd-Günter Voss. 2010. The Working Customer: An Emerg- ing New Type of Consumer. Journal Psychologie des Alltagshandelns/Psychology of Everyday Activity, 3: 2‒10. Ritzer, George. 1983. Sociological theory. New York: Knopf. Ritzer, George and Nathan Jurgenson. 2010. Production, Consumption, Pro- sumption: The Nature of Capitalism in the Age of the Digital ‘Prosumer’. Journal of Consumer Culture, 10: 13‒36. 99 Ritzer, George. 2014. Prosumption: Evolution, Revolution, or Eternal Return of the Same? Journal of Consumer Culture, (14)1: 3‒24. Ritzer, George. 2015. Prosumer Capitalism. Sociological Quarterly, 56: 413‒445. Ritzer, George. 2017. ‘From Production and Consumption to Prosumption: A Personal Journey and Its Larger Context.’ In Margit Keller, Margit Keller, Bente Halkier, Terhi-Anna Wilska, and Monica Truninger (Eds.), Routledge Handbook of Consumption. London: Routledge. Ritzer, George. 2018. The McDonaldization of Society Into the Digital Age. Los Angeles: Sage. Ritzer, George and Degli Esposti Piergiorgio. 2020. The increasing centrality of prosumption in the digital capitalist economy. OZS Oesterreichische Zeitschrift Fuer Soziologie, 2020, 45, pp. 351– 369. doi.org/10.1007/s11614-020-00422-z. Scheiber, Noam. 2019. Inside an Amazon Warehouse, Robots’ Ways Rub off on Humans. New York Times. 3 July, 2019, https://www.nytimes.com/2019/07/03/ business/economy/amazon-warehouse-labor-robots.html (accessed October 17, 2020). Schmidt, Florian A. 2017. Digital Labour Markets in the Platform Economy: Mapping the Political Challenges of Crowd Work and Gig Work. Bonn: Frie- drich-Ebert-Stiftung. http://library.fes.de/pdf-files/wiso/13164.pdf (accessed May 4, 2020). Tapscott, Don and Anthony D. Williams. 2008. Wikinomics: How Mass Col- laboration Changes Everything. Journal of Communication, (58)2: 402‒403. Toffler, Alvin and Heidi Toffler. 2006. Revolutionary Wealth. New York: Doubleday. Toffler, Alvin. 1980. The Third Wave. New York: William Morrow and Company. van Dijck, José, Martijn de Waal, and Thomas Poell. 2018. The Platform Society: Public Values in a Connective World. New York: Oxford University Press. Vargo, Stephen L. and Robert F. Lusch. 2004. Evolving to a New Dominant Logic for Marketing. The Journal of Marketing, 68: 1‒17. Zelizer, Viviana. 1983. Morals and Markets. Princeton, N.J.: Princeton University Press. Zraick, Karen. 2019. Sweetgreen Scraps Its Cashless Policy as Criticism Grows. New York Times. 25  April, 2019, https://www.nytimes.com/2019/04/25/ business/cashless-stores-sweetgreen-amazon-go.html (accessed July 15, 2020). Zuboff, Shoshana and Karin Schwandt. 2019. The Age of Surveillance Capitalism. London: Profile Books. 100 5 Varieties of Internet Platforms and their Transformative Capacity Ulrich Dolata 5.1 Introduction Internet-based digital platforms have grown rapidly in stature and size over the 2010s. Among these are the many search, networking and messaging, advertising, commercial, booking and media platforms that comprise the key socio-technical infrastructures and institutions of today’s Internet. Building on the debates about a digital economy and the emergence of platform capitalism, this chapter focuses on the structures, functions, and reach of commercial platforms on the Internet as well as the interaction of concentration and competitive dynamics in platform markets. From an economic point of view, online digital platforms exert a radical restructuring pressure in particular on already existing economic sectors. However, the truly novel aspect that sets these platforms apart and distinguishes them from their predecessors is that they go well beyond the structuring of economic contexts and reach deep into society: through them, large segments of the private life and public exchange on the Internet are privately organized, curated and commodified. The the- sis put forward here is that the major Internet platforms such as Facebook, Instagram and YouTube organize, observe and regulate significant parts of social exchange on the Internet today—via self-created social rules that can be read in their terms and conditions and community standards, and whose implementation is primarily algorithmic. The private-sector operators of the major platforms have thus taken over essential social structuring and regulation functions on the Internet and have created a parallel social world that has so far neither been democratically legitimized nor democratically controlled. 101 5.2 Typology and Definition From an empirical point of view, the numerous platforms on the Internet differ significantly from one another, calling for a typifying view. The following characteristics of platforms can be distinguished from one another based on their range of services: › Search platforms that are provided by Google as a monopoly or that are oriented towards Google; › Networking and messaging platforms, such as Facebook (with WhatsApp and Instagram), Twitter, or Snapchat; › Media platforms, such as YouTube, Netflix, Apple, or Spotify; › Trading platforms, such as Amazon, Alibaba, eBay, or Zalando; › Booking or service platforms, for example, in the area of ride-hailing services (Uber, Lyft), travel and accommodation booking (Airbnb, Ex- pedia, Booking.com), or dating services (Match, Parship); › Cloud platforms, such as Amazon Web Services, Google Cloud, or Mi- crosoft Azure Cloud, to which individual users, business customers, as well as government institutions outsource their data and the processing thereof; › Crowdsourcing and crowdfunding platforms, such as Amazon Me- chanical Turk, TaskRabbit (a part of the IKEA Group), Kickstarter, or Indiegogo, which serve as hubs for the competition-based awarding of work orders or in order to finance projects. Overall, these platforms can be seen to comprise digital, data-based, and algorithmically structuring socio-technical infrastructures that facilitate the exchange of information, the configuration of communication, the organi- zation of work and markets, the provision of a broad spectrum of services, and the distribution of digital and non-digital products (Kenney and Zysman 2016; Srnicek 2017, 43‒48). As technical infrastructures, they are based on new possibilities for collecting and processing large amounts of data; the comprehensive digital networkability not only of media, information, and communication but also of material things and production structures; and the sorting and coordination of these processes through learning algorithms (Gillespie 2014; Gillespie 2016). As socio-economic units, platforms are not crowd- or sharing-based (Sundararajan 2016)—even if their success (or failure) depends heavily on the number of users and on their personal contributions, communications, ratings, and preferences—but are installed, organized, and controlled top-down by profit-oriented companies. 102 5.3 Economic Features and Reach Beyond this lowest common denominator, the field becomes quite hetero- geneous. Indeed, the various Internet platforms differ significantly from one another not only in terms of classic economic indicators, such as their turnover, profit, market capitalization, or employment, but also in terms of their economic or social reach and significance (Dolata 2018a; Dolata 2019; Van Dijck et al. 2018, 12‒22). The leading Internet groups Google, Amazon, Facebook, and Apple offer a broad spectrum of coordinated and networked services and businesses, which they have developed into extensive socio-technical ecosystems that extend far beyond their traditional field of activity. Google has long ceased to be just a search engine. It owns YouTube, by far the largest video channel on the web; Google Play, the largest app store next to Apple, offering media content of all kinds; Gmail, the leading email service; Google Maps, the most widely-used map service; and Android, the leading operating system for mobile devices. Finally, Google is one of the largest providers of cloud services next to Amazon and Microsoft. Facebook, for its part, together with its subsidiaries WhatsApp and Instagram, is the undisputed leader in social networking and messaging. Over the past decade, Apple and Amazon have also distinguished themselves as full-service providers of a broad range of services and media content, some of which they now produce themselves. The private-sector regulation of the Internet is essentially carried out via these broadly-based platforms that reach deep into the web and whose services are systematically accessed not only by individual users but also by numerous companies, media producers, government institutions, or other platform companies (Barwise and Watkins 2018). In contrast, the countless smaller Internet companies offer more specific services on their platforms. As a rule, these are singular and specialized con- sumer or service offerings that are either purely consumer-oriented, such as ride-hailing services, travel bookings, room referrals, video-on-demand services, and shopping portals, or, like Twitter or Snapchat, communication-oriented. They offer a limited range of services and can generally be assigned to tradi- tional economic sectors, some of which are radically realigned by the activities of the new players. Uber, for example, has brought new momentum to the markets for ride-hailing services, and Airbnb has brought new dynamics to the network-based brokerage of accommodations. Over the past decade, Netflix has developed from a classic video rental service to the world’s leading film streaming service, with its own film productions. However, many of these platforms are dependent on the infrastructure of the big Internet companies. 103 For example, Netflix and Spotify run entirely on the servers of Amazon Web Services and Google Cloud, respectively; and Airbnb and many others inte- grate the Google Maps’ geographical navigation service into their offerings. Looking more closely at the various offerings of the platform operators and the business models behind them, a notable paradox emerges: In the 2010s, the Internet has brought forth many new commercial service offer- ings alongside—on the non-commercial side of the markets—very attractive search, information and user-friendly media, networking, and messaging platforms. However, this has neither led to the emergence of fundamentally new economic sectors nor to the establishment of new and previously un- known business models. The commercial platforms on the Internet do not constitute a new demar- cated economic sector—being much too heterogeneous for that—but offer services that can be economically assigned to traditional sectors and markets such as commerce, advertising, and various service sectors and are empirically difficult to grasp under the general umbrella term ‘digital economy’ (Barefoot et al. 2018). Amazon has revolutionized commerce, and Google and Face- book have expanded the classic advertising markets with the fast-growing segment of online advertising. Uber, applying new transportation concepts, and Airbnb brought new momentum to the markets for ride-hailing services and the web-based booking of accommodations. Netflix has evolved from a classic video rental service to the world’s leading movie streaming service over the past decade. Together with newcomers such as Tesla, Google and Apple have also put manufacturers and suppliers in classic industries, such as the automotive industry, under massive pressure with their activities. All of these developments involved more or less radical readjustments in long-standing sectors and markets, which put the respective established players under massive pressure to adapt and change. However, new major industries, such as the electrical and chemical industry in the early 20th century or the computer, software and semiconductor industries in the 1980s, have not emerged as a result of the activities of these platform companies (Mowery and Nelson 1999). It is also remarkable that the economic and employment effects which the spread of these platforms has entailed have so far remained rather modest. An empirical study by the Bureau of Economic Analysis at the US Department of Commerce estimated that the total number of people employed in the digital economy, which includes the entire information and communications technology industry, contributed only 3.9 percent to total employment in the United States in 2016. The share of commercial Internet platforms in total employment was less than 1 percent, in other words, significantly lower even (Barefoot et al. 2018). Moreover, a study by the International Monetary 104 Fund to measure the macroeconomic effects of the digital economy comes to the conclusion, for the United States, that online platforms and services contributed only 1.5 percent to the US gross domestic product (GDP) in 2015 (International Monetary Fund 2018). Hence, the transformation of the economy towards a platform capitalism or a digital platform economy seems to be still a long ways off. 5.4 Concentration, Market Power and Patterns of Competition It is evident that by now a handful of leading companies have emerged that, with their sprawling ecosystems, dominate key cornerstones of both the social usage and the commercial business on the Internet. Search engines, networking and messaging services, app stores, media services, cloud services, Internet advertising and commerce, and operating systems for mobile Inter- net access—in all of these areas, the four major Internet companies Google, Facebook, Amazon and Apple are the internationally leading players, albeit each in different constellations. Unlike their potential competitors, they are all highly profitable and are among the companies with the highest research and development spending in the world. These concentration processes can be attributed, for one, to the interaction between direct and indirect network effects, which is typical for platform mar- kets (Rochet and Tirole 2003; Haucap and Stühmeier 2016). The prominent position of Facebook (with WhatsApp and Instagram) in the area of social networking and messaging, the dominance of Google as a search engine or of YouTube as a video channel, and the central position of Amazon as a trading platform are directly based on the high numbers of users of these platforms, which render them particularly attractive to advertisers, retailers, or product providers. This fosters concentration processes and makes it difficult for alternative offerings to participate as serious competitors. Secondly, the leading Internet companies today have extraordinary and difference-generating financial power that enables them to invest far more than their potential competitors in expanding their own services and technical and logistical infrastructures (such as server architectures, data collection, and evaluation technologies), in the quality of their search algorithms and in the technical integration of their extensive ecosystems, or, as in the case of Amazon, in the group’s own order, logistics and storage systems. That same financial power also enables them to buy interesting external know-how, to secure their domains of business, to penetrate new business areas via some- 105 times very costly acquisitions of other companies, or to eliminate potential competitors by buying them up early on in the game (Nadler and Cicilline 2020, 36‒131, 406‒450). All of this too allows the leading companies to stand out against their competitors, provided they do not make any serious strategic mistakes, and to establish high entry barriers for newcomers seeking to get a foot in the door of segments and markets already occupied by themselves. Although network effects and classical economic concentration processes systematically favour the emergence of dominant companies, they generally do not lead to the formation of monopolies and the complete or extensive elimination of competition and competitors. Concentration tendencies on platform markets, regardless of where they occur, are accompanied by fierce competition. This is not surprising when looking at the smaller Internet companies like Airbnb, Uber, Spotify, or Netflix. Each and every one of them has to contend with intense competition not only from their immediate competitors but also from the leading Internet companies, media companies, and companies from traditional sectors of the economy, who are challenged by them. The still unconsolidated markets that serve these companies are characterized by a competition that is fierce, complex, and highly volatile (Dolata 2018b). It should be noted, however, that the leading Internet companies do not operate in a competition-free zone either. In their business areas, they have not secured monopolies, even if they are dominant players. The highly concentrated market for Internet advertising, where smaller companies or newcomers no longer have a reasonable chance of securing any significant revenue opportunities today, is characterized by a duopolistic competition between Google and Facebook. Internet advertising also competes with other advertising media such as TV and magazines. Online commerce is in the hands of Amazon, by far the largest retailer on the Internet. That said, online commerce still comprises only the smaller segment of the retail sector as a whole, which is dominated by large retailers such as Wal-Mart. Even Apple is not a monopolist in the market for multimedia devices, instead facing strong competitive pressure mainly from Asian suppliers such as Samsung Electronics or Huawei (Nadler and Cicilline 2020, 77‒131). In addition, the drive of Internet companies to expand beyond their traditional business fields regularly leads to new competitive dynamics both among one another and with established media, entertainment electronics, and technology groups. Google, Amazon, Apple, and Facebook have gradually developed into Internet-based media groups and have established themselves as full-line providers of a wide range of commercial services and media content. They have a broad portfolio of media offerings that allows them to 106 penetrate the domains of classic media groups (film, music, book publishers) and established game providers (such as Microsoft, Sony and Nintendo) as well as network-based distribution and streaming companies (such as Netflix or Spotify). Moreover, Amazon, Apple, Google, and Microsoft have since become major competitors of storage space, computer capacity, and cloud services. Finally, the Internet companies regularly compete for supremacy in new technical trends such as virtual reality, machine learning, or image and speech recognition among themselves (Dolata 2018a; Dolata 2019). Thus, even within its top tier, the commercial Internet is characterized by strong concentration tendencies as well as by intense competition in all its essential segments, which repeatedly challenges the dominance of individual companies. The extraordinary volatility of services, markets, technologies, and users is driving these companies to permanently defend and renew their competitive and market power. The failure to properly evaluate a new trend, ignoring user preferences or a bigger-than-usual product flop can be enough to put them in a fairly difficult situation (Mellahi and Wilkinson 2004). To avoid this from happening, all platform-based companies are forced to be permanently adaptable; in other words, they must be able to anticipate, take in, and integrate, early on and continuously, new technological developments and socio-economic trends and to transform them into attractive commercial offers (Dolata 2013, 56‒93). However, the competitive pressure to which the leading companies are exposed is no longer generated primarily by commercially oriented newcomers but instead above all by their direct competitors. That pressure manifests as an oligopolistic competition between the leading companies, which is carried out mainly through aggressive innovation and expansion strategies into new areas. Under these oligopolistic conditions, the only chance smaller platform companies have of becoming significant (co-)players is to manage to occupy new, still unconsolidated commercial fields that are not yet on the radar of established companies. 5.5 What’s New? Much of what characterizes commercial platforms on the Internet econom- ically seems to be not really new and disruptive. The platform companies operate with a very narrow and far from spectacular set of business mod- els—advertising, commerce, subscriptions, or booking fees. They do not themselves constitute a new industry sector of significance, as was the case at the beginning of digitization in the 1980s with the emergence of the PC, 107 software, and semiconductor industries. In addition, it seems that so far this part of the (digital) economy is of extremely low macroeconomic significance. So: What’s new and disruptive? To answer this question, we have to bear in mind that the aforementioned economic features alone do not adequately reflect both the considerable influence which the leading Internet groups wield on the readjustment of economic structures and processes and the extraordinary social and socio-po- litical clout that they have attained. The rapid spread of commercial Internet platforms over the past two decades has not only triggered massive upheavals and induced substantial restructuring processes in a number of established economic sectors (e.g. retail, advertising markets, media, various service and industry sectors) but has also allowed a number of Internet companies to establish themselves as rule-setting coordinators of corporately owned and internationally oriented markets. In addition, large parts of the social exchange on the web, from private communication and personal self-presentation to the most diverse kinds of public spheres, are now bundled, evaluated, and curated by a few commercially operated platforms. The private platforms’ roles as organizers of markets and curators of so- cial contexts are, along with the commodification of user behaviour (Zuboff 2019), the essential characteristics that make them a disruptive force and enable them to act as central regulatory bodies in today’s Internet. 5.6 Organization of Markets The leading Internet companies have long since been much more than dominant economic actors who compete with other market players. In ad- dition, they are operating, coordinating, and controlling their own markets as well. In these privately owned and online-mediated markets, the Internet companies assume the rule-setting role of market coordinators: they do not act merely as intermediaries who simply make market transactions of third parties technically possible, but rather structure, regulate, and monitor the activities of all market participants. This affects some of the major platforms of the leading Internet groups. Indeed, Amazon maintains the largest trading platform for third-party pro- viders on the Internet, Amazon Marketplace, which by now generates higher sales than the corporation’s own online retail business (Khan 2018). Google operates YouTube, a central media platform on the web, and organizes the framework conditions and monetization opportunities for YouTuber and In- fluencer as well as professional media producers through its YouTube Partner 108 Program (Burgess and Green 2018). Apple, Google, and Amazon also have large app stores where software developers compete for commercial attention, based on guidelines and commission models set by the market coordinators (Dolata and Schrape 2014). As a result, privately regulated and socio-technically constituted market regimes have taken shape on the Internet that clearly stand out from other markets. They are neither primarily organized, regulated, or guaranteed by the state nor do they constitute themselves through the self-organized and deliberative interaction of various non-state actors (Aspers 2011, 148‒168; Ahrne et al. 2015). Instead, they are installed, operated, and controlled by individual companies. The platform operators act neither as competing market participants nor as neutral intermediaries, but rather as rule-setting and regulatory actors who endow themselves with far-reaching authority and powers of intervention and who thus assume essential functions that are prerequisites for the acceptance, functionality, and reliability of the market. Further, the technical infrastructures provided by the platform operators are not neutral architectures through which connections are merely established. Instead, through the rules inscribed in them, they form these markets’ in- stitutional foundation, the basis that guides actions and structures processes and to which providers, consumers, and users must orient themselves if they wish to play a part. Plans to establish platform-specific private currencies go a significant step further. With this, the privatization of market regimes described above could be extended to include the much more far-reaching prospect of private-sector regulation of macroeconomic interrelationships. Eventually, sovereign tasks, previously performed primarily by democratically legitimized and politically independent institutions, could be, at least partially, delegated to private companies or consortia. This could concern, for example, the regulation of money supply, interest rate policy, and the safeguarding of price level stability or banking supervision, which have so far been the domain of central banks. Such plans are most advanced at Facebook. In mid-2019, with the Libra project, the social media company presented not only an initial concept for a digital currency but also an appropriate regulatory and institutional frame- work (Schmeling 2019; Taskinsoy 2019; Mai 2019). The core organization slated to spearhead this project was the Libra Association, a consortium of Internet companies, payment providers, and other organizations, designed as a private-sector counterpart and parallel structure to the central banks. This body was intended to not only be responsible for the design and enforce- ment of Libra rules and the technical infrastructure of the digital currency but also for managing the Libra reserve, create Libra money and control 109 the money supply, monitor payment channels and admit new Libra traders (Libra Association 2019). Although these plans have since been scaled back following massive political pressure (and the Libra Association was renamed to Diem Association), their basic direction is clearly recognizable. Their general direction of impact was the bid to relativize the importance of central banks and governments in a central area of macroeconomic management and to supplement or replace these with private-sector forms of macroeconomic regulation. In this sense, the original plan comprised the takeover of quasi-sovereign economic regulatory tasks by the private sector, which, as we will see below, are substantially expanded by the assumption of quasi-sovereign social structuring and curating tasks. 5.7 Curation of Sociality In addition to organizing and regulating markets, the platforms—in partic- ular the widely built-out and networked ecosystems of the leading compa- nies—have taken over essential social ordering and regulatory functions on the Internet, which are summarized here as curation of social relationships and social behaviour. Through their numerous services and offerings, these platforms filter information and communication processes, shape individual behaviour and organizational action, and structure social relationships and public spheres—and do so in a far more comprehensive manner than even large media corporations have ever been able to do (Couldry and Hepp 2016, 34‒56). While media corporations remain embedded in society and in its institutional structure as powerful opinion-forming actors with a limited reach, the large platforms, with their own rule-setting, structuring, selection, monitoring, and sanctioning activities, constitute no less than the institutional foundations of a private-sector sociality on the Internet, which have, over the past two decades, evolved largely decoupled from democratic institutions and state influence (Dolata 2020). The basis of curation is formed by binding and sanctionable social rules. They are expressed in the general terms and conditions of the companies and, above all, in community standards (Facebook), guidelines and rules (YouTube; Twitter), in which the platform operators formulate in detail what they con- sider to be politically unacceptable, offensive, obscene, erotic or pornographic, or glorifying violence and terrorism. These guidelines that form the basis of social curation are largely translated into technical instructions, structurings, sortings, and rankings, which I refer to as technically mediated curation. 110 First, this manifests as a technically mediated structuring and design of social action frameworks that both enable and channel the activities of a diverse range of users. This includes the given user interfaces and default settings of the platforms, which have an action-structuring effect by enabling certain activities and excluding or impeding others. The numerous features embedded in the platforms (such as Facebook’s Reactions or Twitter’s Trending button) can also be summarized as action-orienting and opinion-forming structural elements inscribed in technology. In addition, Application Programming Interfaces (APIs) are used to integrate the web presences of countless third parties into the platforms’ scope of action and to establish extensive links between the platforms and external websites, other platforms and apps (Van Dijck 2020; Nieborg and Helmond 2019; Helmond 2015; Gerlitz and Helmond 2013). Secondly, these structure-building effects of technology are supplemented by approaches to a technically mediated institutionalization of social rules and regulation of social processes, which is implemented primarily through the use of algorithms. Algorithms translate the social rules and norms that are valid on the platforms into technical instructions; monitor and sanction participants’ activities; decide what is important and what is not, according to social relevance criteria inscribed in them; select, aggregate, and rank informa- tion, news, videos, or photos on this basis; structure private information and communication processes as well as public discourses; and constitute public spheres and communities that would not exist without them. With all this, algorithms essentially become the nucleus of a technically mediated framing, control, and curation of social action on platforms (Gillespie 2014; Gillespie 2016; Just and Latzer 2017; Yeung 2018; Katzenbach and Ulbricht 2019). Another major step in the curation of sociality is the establishment of a corporate-owned oversight body at Facebook, responsible for monitoring, moderating, and evaluating content on the platform. The Oversight Board, active since 2020, staffed with external experts and financed by the company, seeks not only to monitor and further develop the implementation of the social rules laid down in the Community Standards but also has the authority to judge disputed content and, if necessary, have it removed from the platform (Harris 2020). This regulatory body with a quasi-sovereign function is set up as a kind of constitutional court and supervisory committee, albeit without the democratic legitimacy of such bodies or the ability to exert influence on fundamental corporate decisions. As a result of the combination of these factors, especially the leading Internet groups are now far more than infrastructure providers that provide connectivity; media groups that have a broad portfolio of their own media offerings; or advertising, retail, hardware, and service companies that continue 111 to generate the majority of their revenues and profits with their traditional businesses. The few large platforms that today both enable and shape large parts of private and public life on the Internet can be understood as differ- entiated societal systems with a distinct institutional foundation, which the companies as platform operators structure and control to a considerable extent and by means of their own rules, regulations, and committees—right up to the assumption of quasi-sovereign tasks by the companies that, hitherto reserved for state authorities, so far largely skirt democratic legitimation and control. 5.8 Conclusion and Outlook: Political Regulation of Platforms? The economic but above all social structuring and regulatory power that the leading Internet companies have attained with their platforms is what I refer to as regulation by platforms: the intentional structuring and regulating not only of economic markets but also, and in a much more comprehensive way, of larger societal relations and processes, carried out by Internet companies as platform operators and aligned with their economic exploitation interests. Against this background, I will conclude by briefly discussing the scope for a political regulation of platforms. In fact, the second half of the 2010s has seen an increase in government efforts to achieve political regulation and control of the major platforms. In Europe, since the mid-2010s, such activities have been concentrated in two main areas of action. A first line of activities brought forward above all by the European Com- mission has attempted to limit economic market power. The Commission has pursued a series of infringements of EU antitrust law by Internet com- panies and has repeatedly imposed heavy fines, especially on Google and on Facebook, among others, for an abuse of their dominant position in online advertising, with search engines or through the mobile operating system An- droid (Viscusi et al. 2018, 404‒419; Haucap and Stühmeier 2016; European Commission 2019). A second line of activities has focused on legal and regulatory interven- tions in the regulatory sovereignty of platforms—for example, in the form of the European General Data Protection Regulation (GDPR); the ‘right to be forgotten’ on the Internet, introduced by the European Court of Justice, in a landmark decision; or the German Network Enforcement Act (NetzDG), which obliges the providers of leading social networks such as Facebook, YouTube, or Twitter to block illegal content in a timely manner or to remove 112 it from their platforms and to report on it on a regular basis (Schulz 2018; Chenou and Radu 2019). However, the scope of these political interventions has so far remained extremely limited. Paradoxically, these attempts by the state to intervene in the social regulatory sovereignty of platform operators have tended to strength- en the regulatory power of the platforms, namely, by delegating sovereign functions of jurisdiction and enforcement to private sector actors and by providing this shift with political legitimacy. Germany’s Network Enforcement Act, for example, has done little to change the fact that companies such as Facebook, Google, or Twitter largely decide for themselves which content they delete and which they do not, and has, indeed, even strengthened the companies in their role as content moderators and as decisive instances of content evaluation or selection. Further, the enforcement of the right to be forgotten has also been assigned to the platforms themselves, which have thus become more integrated into the legal system and, as private-sector organizations, have been entrusted by the government with quasi-sovereign tasks (Chenou and Radu 2019). Overall, the political regulatory approaches, to date, are not suitable for substantially correcting or controlling the regulatory sovereignty of the platform operators. However, the presentation of a Digital Markets and Services Act by the EU Commission at the end of 2020 (European Commission 2020a; European Commission 2020b) and a lawsuit filed by the US Federal Trade Commission against Facebook, which aims for nothing less than a split-up of the group, show that the question of how the overwhelming power of Internet corporations and their platforms can be limited and more publicly controlled is no longer being considered only in Europe but now also in the United States. In this context, two more far-reaching directions in which considerations about stronger political regulation of Internet corporations should develop are becoming increasingly apparent. These considerations include first the radical unbundling of the widely networked platforms of the Internet corporations—such as the decoupling of YouTube and other platforms from the Google corporation, or the split- ting up of the ecosystem of Facebook, Instagram, and WhatsApp (Nadler and Cicilline 2020, 378‒382). However, such considerations, which would, admittedly, involve a rather brutish dismantling, should be justified less by a limitation of these corporations’ economic market power than by the aim of limiting their extraordinary socio-political structuring and regulatory power. A second direction in which the discussions should go is the setting up of public supervisory and regulatory bodies, for example, at the European and US levels. Controlled by parliament and staffed with recognized and pub- 113 licly appointed experts, these authorities should be set up as democratically legitimate alternatives to corporate supervisory bodies (such as Facebook’s Oversight Board) and be equipped with far-reaching information, control, and sanctioning powers. They could also be tasked to disclose, control, and impose conditions on algorithmic filtering functions, ranking and rating principles, as well as community standards, and the search and selection criteria based upon them (Dolata 2018c). However, even the proposal for public supervisory and regulatory au- thorities would not, if implemented, lead to a private-state co-regulation of platforms on an equal footing—if only because of the extreme informa- tion and knowledge asymmetries of the parties involved. Indeed, political regulators are much less knowledgeable about the extensive socio-technical systems and systemic contexts they are supposed to regulate than those who have developed and now operate these systems. Hence, in this case, too, the responsibility for structuring and regulating economic and social processes on the Internet would remain primarily with the platform operators. But at least then their activities could be regularly evaluated, controlled, and sanctioned by a democratically legitimized body. 5.9 References Ahrne, Göran, Patrik Aspers, and Nils Brunsson. 2015. The Organization of Markets. Organization Studies, 36(1): 7‒27. Aspers, Patrick. 2011. Markets. Cambridge, Malden: Polity Press. Barefoot, Kevin, Dave Curtis, William Jolliff, Jessica R. Nicholson, and Robert Omohundro. 2018. Defining and Measuring the Digital Economy. Working Paper 3/15/2018. Washington D.C.: Bureau of Economic Analysis Barwise, Patrick and Leo Watkins. 2018. ‘The Evolution of Digital Dominance: How and Why We Got to GAFA.’ In Martin Moore and Damian Tambini (Eds.), Digital Dominance. The Power of Google, Amazon, Facebook, and Apple. Oxford: Oxford University Press, 21‒49. Burgess, Jean and Joshua Green. 2018. YouTube. Online Video and Participatory Culture. 2nd ed. Cambridge and Medford: Polity. Chenou, Jean-Marie and Roxana Radu. 2019. The “Right to Be Forgotten”: Negotiating Public and Private Ordering in the European Union. Business & Society, 58(1): 74‒102. Couldry, Nick and Andreas Hepp. 2016. The Mediated Construction of Reality. Cambridge, Malden: Polity. Dolata, Ulrich. 2013. The Transformative Capacity of New Technologies. A Theory of Sociotechnical Change. London and New York: Routledge. 114 Dolata, Ulrich. 2018a. ‘Internet Companies: Market Concentration, Competi- tion and Power.’ In Ulrich Dolata and Jan-Felix Schrape (Eds.), Collectivity and Power on the Internet. A Sociological Perspective. Cham: Springer, 85‒109. Dolata, Ulrich. 2018b. Die zweite Reihe. Neue Konkurrenten für die Internetkon- zerne? SPW – Zeitschrift sozialistische Politik und Wirtschaft, 41(225): 23‒30. Dolata, Ulrich. 2018c. Big Four – Die digitale Allmacht? Blätter für deutsche und internationale Politik, 63(5): 81‒86. Dolata, Ulrich. 2019. Privatization, Curation, Commodification. Commercial Platforms on the Internet. Österreichische Zeitschrift für Soziologie, 44: 181‒197. Dolata, Ulrich. 2020. Internet—Platforms—Regulation. Coordination of Markets and Curation of Sociality. SOI Discussion Paper 2020‒02. Stuttgart: Institute for Social Sciences. Dolata, Ulrich and Jan-Felix Schrape. 2014. App-Economy: Demokratisierung des Software-Marktes? Technikfolgenabschätzung – Theorie und Praxis, 23(2): 76‒80. European Commission. 2019. Antitrust: Commission Fines Google € 1.49 Billion for Abusive Practices in Online Advertising. Brussels: European Commission- Press Release. 20 March, 2019, https://ec.europa.eu/commission/presscorner/ detail/en/IP_19_1770 (accessed March 12, 2021). European Commission. 2020a. Proposal for a Regulation of the European Parliament and of the Council on a Single Market for Digital Services (Digital Services Act) and amending Directive 2000/31/EC. COM(2020)825 final. Brussels: EU Commission. 15 December, 2020, https://ec.europa.eu/info/sites/info/ files/proposal_for_a_regulation_on_a_single_market_for_digital_services. pdf (accessed March 12, 2021). European Commission. 2020b. Proposal for al Regulation of the European Parliament and of the Council on Contestable and Fair Markets in the Digital Sector (Digital Markets Act). COM(2020)842 final. Brussels: EU Commission. 15 December, 2020, https://ec.europa.eu/info/sites/info/files/ proposal-regulation-single-market-digital-services-digital-services-act_en.pdf (accessed March 12, 2021). Gerlitz, Carolin and Anne Helmond. 2013. The Like Economy: Social Buttons and the Data-Intensive Web. New Media & Society, 15(8): 1348‒1365. Gillespie, Tarleton. 2014. ‘The Relevance of Algorithms.’ In Tarleton Gillespie, Pablo Boczkowski, and Kirsten Foot (Eds.), Media Technologies. Essays on Communication, Materiality, and Society. Cambridge: MIT Press, 167‒194. Gillespie, Tarleton. 2016. ‘Regulation of and by Platforms.’ In Jean Burgess, Thomas Poell, and Alice Marwick (Eds.), The SAGE Handbook of Social Media. Los Angeles: Sage, 254‒278. Harris, Brent. 2020. Preparing the Way Forward for Facebook’s Oversight Board. 28 January, 2020, https://about.fb.com/news/2020/01/facebooks-oversight-board/ (accessed March 12, 2021). 115 Haucap, Justus and Torben Stühmeier. 2016. ‘Competition and Antitrust in In- ternet Markets.’ In Johannes M. Bauer and Michael Latzer (Eds.), Economics of the Internet. Cheltenham and Northampton: Edward Elgar, 183‒210. Helmond, Anne. 2015. The Platformization of the Web: Making Web Data Platform Ready. Social Media & Society, 1(2): 1‒11. International Monetary Fund. 2018. Measuring the Digital Economy. Washington D.C.: IMF. Just, Natascha and Michael Latzer. 2017. Governance by Algorithms: Reality Construction by Algorithmic Selection on the Internet. Media, Culture & Society, 39(2): 238‒258. Katzenbach, Christian, and Lena Ulbricht. 2019. Algorithmic Governance. Internet Policy Review, 8(4): 1‒18. Kenney, Martin and John Zysman. 2016. The Rise of the Platform Economy. Issues in Science and Technology, 32(3): 61‒69. Khan, Lina M. 2018. ‘Amazon—An Infrastructure Service and its Challenge to Current Antitrust Law.’ In Martin Moore and Martin Tambini (Eds.), Digital Dominance. The Power of Google, Amazon, Facebook, and Apple. Oxford: Oxford University Press, 98‒129. Libra Association. 2019. An Introduction to Libra. White Paper. Geneva. 23 July, 2019, https://www.diem.com/en-us/white-paper/ ( accessed March 12, 2021). Mai, Heike. 2019. Libra—A Global Challenger in Payments and for Central Banks? EU-Monitor—Digital Economy and Structural Change. 22 July, 2019. Mellahi, Kamel and Adrian Wilkinson. 2004. Organizational Failure: A Critique of Recent Research and a Proposed Integrative Framework. International Journal of Management Reviews, 5/6(1): 21‒41. Mowery, David C. and Richard R. Nelson (Eds.). 1999. Sources of Industrial Leadership. Studies of Seven Industries. Cambridge: Cambridge University Press. Nadler, Jerrold and David N. Cicilline. 2020. Investigation of Competition in Digital Markets. Majority Staff Report and Recommendations. Subcommittee on Antitrust, Commercial and Administrative Law of the Committee on the Judiciary. Washington D.C.: United States House. Nieborg, David B. and Anne Helmond. 2019. The Political Economy of Facebook’s Platformization in the Mobile Ecosystem: Facebook Messenger as a Platform Instance. Media, Culture & Society, 41(1): 196‒218. Rochet, Jean-Charles and Jean Tirole. 2003. Platform Competition in Two-Sided Markets. Journal of the European Economic Association, 1(4): 990‒1029. Schmeling, Maik. 2019. What is Libra? Understanding Facebooks Currency. SAFE Policy Letter No. 76. Frankfurt a.M.: Goethe University—House of Finance. Schulz, Wolfgang. 2018. Regulating Intermediaries to Protect Privacy Online— The Case of the German NetzDG. HIIG Discussion Paper Series 2018‒01. Berlin: Alexander von Humboldt Institute for Internet and Society (HIIG). 116 Srnicek, Nick. 2017. Platform Capitalism. Cambridge and Malden: Polity Press. Sundararajan, Arun. 2016. The Sharing Economy. The End of Employment and the Rise of Crowd-based Capitalism. Cambridge and London: MIT Press. Taskinsoy, John. 2019. Facebook’s Project Libra: Will Libra Sputter Out or Spur Central Banks to Introduce Their Own Unique Cryptocurrency Projects? Sarawak: University Malaysia Sarawak. Van Dijck, José, 2020. Seeing the Forest for the Trees: Visualizing Platformization and its Governance. New Media & Society, 1‒19, DOI10.1177/1461444820940293 (accessed March 12, 2021). Van Dijck, José, Thomas Poell, and Martijn de Waal. 2018. The Platform Society. Public Values in a Connective World. Oxford: Oxford University Press. Viscusi, W. Kip, Joseph E. Harrington Jr., and David E.M. Sappington. 2018. Economics of Regulation and Antitrust. 5th ed. Cambridge: MIT Press. Yeung, Karen. 2018. Algorithmic Regulation: A Critical Interrogation. Regulation & Governance, 12(4): 505‒523. Zuboff, Shoshana. 2019. The Age of Surveillance Capitalism: The Fight for the Future at the New Frontier of Power. London: Profile Books. Part III Platform Work 119 6 From Commodification to Free Labour: The Gendered Effects of Marketplace Platforms on Work Anne Jourdain 6.1 Introduction The rise of marketplace platforms (Airbnb, Craigslist, Upwork, etc.) is en- couraging people to commodify personal goods and services, as profession- als or amateurs. This ‘marketization of everyday life’ (Naulin and Jourdain 2019) is driven by the development of digital technologies lowering the cost of market access, but also by the global economic crisis prompting people in insecure situations to look to any means to earn extra money. The con- temporary values of capitalism and neoliberalism—with the glorification of the post-Fordist figure of the entrepreneur (Adkins and Dever 2016)—also explain the growing trend of commodification. The new digital economy based on web platforms and social media (Facebook, Instagram, Twitter, etc.) bears the promise that, coupled with the ‘sharing economy’ (Botsman and Roger 2011), it will create new business opportunities and new jobs (especially in the form of self-employment). In bringing ‘the romance of entrepreneurialism to the masses’ (Ravenelle 2017, 1), it is supposed to be conducive to both worker empowerment and economic growth. However, most recent studies of the digital economy have shown that web platforms are rather associated with new forms of exploitation (Huws 2003; Scholz 2012), poor-quality risky jobs (Ravenelle 2019), and growing inequalities between workers (Casilli 2019; Schor 2018). In line with these studies, this chapter delves behind the scenes of the digital economy and explores the perverse effects of the marketization of everyday life. In order to explore the gendered effects of the digital economy on work, this chapter is based on an empirical study of the marketplace platform Etsy which bears the promise of a female emancipation. This leading web platform, created in 2005, sells essentially handmade products (bags, clothes, furniture, 120 jewellery, etc.). Sellers set up their own virtual shops on the Etsy website and the platform makes a commission of approximately 5 percent on each trans- action (plus $ 0.20 per displayed item). Today, 3.7 million creators sell their wares on Etsy worldwide. A full 88 percent of them are women. This female phenomenon of commodification is fostered by the ‘new domesticity’ trend, as journalist E. Matchar (2013) portrayed it in her book Homeward Bound. ‘New domesticity’ refers to white middle-class mothers in their thirties who escape traditional companies to work from home and balance work and family constraints. The idea is that by making money from a leisure activity, hand- icrafts, in this case, they can both take care of their family and earn income. This ‘having it all’ ideal (Adkins and Dever 2016) is thereby synonymous with social upgrading and empowerment. Like professional female designers, Etsy’s female hobbyists may also be won over by such narratives, even though they often hold a salaried job alongside their commodified leisure activity. However, as I will show, the promise is somewhat empty since earnings made from the platform are extremely low, for both hobbyists and professionals attracted by the platform’s tagline, ‘Turn your hobby into a business’. Commodification on marketplace platforms such as Etsy has seen the rise of new forms of activity, which could be described as ‘work’ or even ‘extra work’. The development of web platforms has blurred the boundaries between work and non-work (leisure, domesticity, and free time), prompting a recon- sideration of the nature of work (Flichy 2017). Many academics (Cardon and Casilli 2015; Casilli 2019; Scholz 2012) drawing on a Marxist reading grid use the concept of ‘digital labour’ to define the creation of digital con- tent and data by individuals (on web devices such as Facebook and Google), whereas this activity is often individually perceived as recreational. Given that most of the economic value is appropriated by the digital platforms, ‘digital labour’ is associated with ‘free labour’ (Terranova 2000). Although ‘free labour’ appears to be a recent theme in digital media studies, it was ac- tually conceptualized forty years ago by feminist gender studies of ‘domestic work’ (Jarrett 2015; Simonet 2018). By labelling as ‘work’ that which was not previously understood as work, i.e. women’s domestic activities, feminists criticized the invisibility and exploitation of women’s actions in capitalist, patriarchal societies (Delphy and Leonard 1992; Kaplan Daniels 1987). In so doing, they proposed new ways to analyse work. Drawing on both digital media studies and gender studies, this chapter identifies two forms of ‘extra work’ performed by Etsy sellers: digital labour and domestic work. Digital labour refers to the great deal of time Etsy sellers spend managing their online shop on the platform, posting good pictures of their products, choosing suitable tags for their items, posting about their 121 new creations on Facebook, Instagram, Pinterest, and so on.  Domestic work traditionally means housekeeping and childcare: the women who set up a craft business generally take on more household chores, which widens in-household gender inequalities. As I will show, these two forms of work can be conceptualized as ‘free labour’ (Simonet 2018; Terranova 2000) in that they are mostly invisible and generate economic value that does not entirely benefit the Etsy sellers themselves. This chapter finally examines a paradox: whereas commodification aims to earn extra money, it also generates forms of extra work that are not monetized and not socially recognized, in other words, free labour. What specific forms does this free labour take in the case of Etsy? What economic value does it generate? How is this value captured and by whom? Why do commodifiers engage in such free labour? My empirical study of Etsy addresses all these questions. My hypothesis is that the social logics identified by this case study can be generalized to other kinds of commodification on marketplace platforms. This chapter, therefore, undertakes to analyse work transformations in contemporary societies to gain a clearer picture of the future of work. My empirical research on Etsy focuses on French sellers and is based on mixed methods.1 First, I gathered statistical data on Etsy sellers. I used web scraping to collect information on 14,415 French Etsy online shops active in 2017, and 19,715 actives in 2019, 88 percent of which were set up by women. Second, I conducted semi-structured interviews with nineteen Etsy sellers (sixteen women and three men) from 2016 to 2021. I diversified the interviewees’ profiles in terms of age, gender, location, visibility on the web, and type of handicrafts. Third, I conducted ethnographic observations of craft fairs such as Etsy Made in France in Paris. I met and had informal discussions with many Etsy sellers on these occasions. 1 This research is part of two collective projects: the CAPLA project, financed by the French ANR, entitled ‘Workers on tap. The social impact of platform capitalism’ coordinated by Sarah Abdelnour (Paris-Dauphine University, IRISSO) from 2016 to 2021; and the Marketization of Everyday Life project, financed by the Independent Social Research Foundation, coordinated by Sidonie Naulin (Sciences Po Grenoble, PACTE) and me from 2016 to 2019. This project also benefited from collaboration with the Governance Analytics team led by Bruno Chaves Ferreira (Paris-Dauphine University, PSL). 122 6.2 From Platform Promise to Economic and Social Reality Many recent studies of platform capitalism have pointed up the discrepancy between the ‘myth’ promoted by digital platforms and the general ‘disenchant- ment’ felt by their workers (Abdelnour and Méda 2019). Etsy is no exception. Most of its sellers—either hobbyists or (aspiring) professionals—do not earn as much as they initially expected to from the platform. 6.2.1 Small Amounts of Extra Money Far from the platform tagline ‘turn your hobby into a business’ and the me- dia spotlight on success stories concerning Etsy sellers who have managed to earn a living from their craft business, the vast majority of Etsy sellers only earn small amounts of money from the platform. The reconstituted average turnover estimated from my quantitative data on the French online shops is € 91 a month, with 50 percent of sellers earning less than € 10 a month. Income inequalities are particularly high, and, as with many other web plat- forms, Etsy presents a ‘long tail’ (Anderson 2006): 10 percent of Etsy sellers make 88 percent of the platform’s sales. Only 1 percent earn the equivalent of the minimum wage. So only a happy few do a profitable business out of the platform. Many Etsy sellers do not actually see their commodification activity as a true business. They consider themselves to be amateurs, especially when they hold a salaried job in an area that has nothing to do with crafts. Usually middle or upper class and highly educated, they do not really need the extra income from their commodification. When money is actually earned from Etsy, the small amounts are usually reinvested in the activity (Zelizer 1994). However, even as hobbyists, many amateurs do not earn as much money as they expected to when they opened their online shop on Etsy. Some Etsy sellers set out to make a living from their activity. Unlike hob- byists, they often register as self-employed when developing a craft business. Although they make more money on the whole than amateurs, they generally do not manage to earn a living from their online activity. They consequently seek to develop profitable sidelines, such as DIY/craft classes, on the back of the popularity of their online showcase and to diversify their points of sale (craft fairs, decoration shops, etc.). The modest amounts of money made from online marketization can be frustrating, particularly for aspiring professionals. In actual fact, the reality is a far cry from the success stories presented in the media and on platforms like Etsy, which promotes successful ‘featured 123 shops’ on its blog. At first, aspiring professionals regard the happy few who earn a wage from their online activity as proof that their professional dream can come true. As time passes, some of them start to feel bitter about their modest earnings, having hoped they would sell more easily and make more money on the Internet. Their resentment is sometimes levelled at the web platform that has not kept its marketization promise. For example, Elisabeth, who left her job as an IT engineer to earn a living as a seamstress, has only sold a handful of her textile products on the Internet: I’ve actually realized that if we don’t help [the platform] make money, because we don’t sell a lot or we don’t attract new people to the platform or because we don’t buy advertising, we get a poor ranking. It’s a vicious circle. I’m disappointed in the platform. I find it hard. We’re left to fend for ourselves because we’re completely invisible. And we don’t make many sales. (Elisabeth, forty-two years old) Over the years, those who persevere with their business ambitions despite low profits are those who can rely on other economic resources such as a partner’s substantial, steady income, a family inheritance or a side job. The general meagreness of Etsy incomes might explain the high turnover on the platform. Two web scrapings were conducted to compare active French online shops on Etsy in 2017 (14,415 shops) and in 2019 (19,715 shops). I found that 65 percent of the 2017 shops were no longer active on Etsy two years later, i.e. 65 percent of the 2017 Etsy sellers had not paid the $ 0.20 to continue displaying their wares in their online shop in 2019. Not sur- prisingly, Etsy metrics reveal that all these shops pulling out were less active (fewer ‘favourite shops’, fewer stars, etc.) and less profitable (lower incomes) than the others. Their owners may have been hobbyists trying to sell on Etsy ‘for fun’ or aspiring professionals disappointed by their poor earnings on the platform (I met some myself ). A small minority representing the 1 percent happy few (described in more detail in part 6.2) might have used Etsy as a springboard to launch their creative business and build a brand reputation on the Internet. As they became successful and attracted a regular customer base, they would then have closed their online shop on the platform to make it on their own free of Etsy and its transaction costs. 6.2.2 A Gender Gap Among Sellers Whereas Etsy and its community claim to further female empowerment via self-employment and entrepreneurship, statistics reveal gender disparities even on this platform. Men represent just 12 percent of Etsy sellers, but their 124 average monthly turnover on Etsy is € 150 compared with € 84 for women. Such gender gap can be found in other platforms like Upwork (Foong et al. 2018). But in the case of Etsy, it is all the more surprising as this platform is associated with women’s emancipation. In the female field of handicrafts, just like in many other sectors, men are therefore more successful than women. This gender gap among Etsy sellers can partly be explained by the fact that the men and the women do not have the same purposes when they create an online shop on the platform. The three interviews with the male sellers suggest that they approach the marketization of leisure activities from a different angle to the women. The commodification of their activity is more frequently driven by professional ambition and business considerations. Unlike the women, the men are quick to alter their goods and services to make them more profitable. For example, Jonathan, a 41-year-old artist living with his wife and daughter, developed mini-design collections on Etsy alongside his one-off contemporary works of art, because they ‘need to eat too’. The necessity to make ends meet is consubstantial with the ‘breadwinner’ male figure and concerns fewer female sellers, either because they have a primary job next to Etsy or because their male partner provides for them and their family. The hypothesis of a more professionalised orientation for male businesses appears to be borne out by the fact that men are overrepresented among the 1 per- cent of Etsy sellers who earn more than the minimum wage: they represent 24 percent of the happy few whereas they account for just 12 percent of all sellers. Besides, gendered economic differences are smaller among these high earners: the average monthly turnover is € 2,801 for these men (n = 29) and € 2,835 for these women (n = 96). Success among this small, professionalised elite seems to be equally shared by the men and the women. However, the overall gender gap cannot entirely be explained by a female inclination for amateurism and a male preference for professionalism. First- ly, the database only contains active shops and excludes all the ghost shops without sales (except for the few ones which were created in the last four months). The analysis thus ignores mere amateurs who do not intend to sell on the platform. Secondly, the subsample of 125 successful individuals is too small to ensure the robustness of the result of a relative gender equality among the high earners. Moreover, as my qualitative data show, those who intend to earn a living on Etsy are far more numerous than the 1 percent who really manages to earn the equivalent of the minimum wage. Indeed, many women, like Elisabeth (see part 6.1.1), are eager to set up a real business on Etsy but do not make as much as they expected. Finally, the demand side must also be studied to consider the question of gender inequalities fos- tered by the platform. As a matter of fact, the relative success of men in the 125 marketization of female leisure activities may be associated with traditional gender inequalities in the professional handicrafts field, i.e. arts and crafts, where the most profitable businesses are run by men (Jourdain 2014). It may tie in with persistent gender stereotypes that equate men with art and talent, while women are relegated to domesticity and ‘inconsequential’ activities. Such stereotypes would enable men to set higher prices than women. Indeed, on Etsy, the men’s average selling price is € 151 as opposed to € 45 for the women, and such difference remains among the high earners. Consequently, multiple factors account for the objective gender gap on Etsy. The gendered differentiation in the types of (free) work that the platform demands, as analysed below, is another one of them. 6.3 First Form of Free Labour: Digital Labour Marketplace platforms like Etsy use specific algorithms and metrics, which organize Internet users’ actions and the capitalist societies in which they live (Cardon 2015). Buyers use them when looking for certain items on search engines such as Google and directly on marketplace platforms. Sellers, therefore, need to adjust to this digital reality and build their reputation, or ‘e-reputation’ (Beuscart et al. 2015), to rise up the rankings, increase their visibility, and sell more. This race for visibility implies developing different types of online activity. 6.3.1 In the Name of an e-Reputation: The Different Types of Online Activity Etsy sellers do not only craft bags, jewellery, and mugs, they also spend a great deal of time online to sell their products. Commodifying a leisure ac- tivity calls for ‘market work’ (Cochoy and Dubuisson-Quellier 2013), which consists of defining market positioning, setting prices, building a ‘brand’ by creating and maintaining a positive reputation, promoting the products, managing interactions with customers, understanding the trade legislation, and so forth. On marketplace platforms, a large part of this ‘market work’ is an online activity. On the Etsy website, once sellers have opened an online shop, which takes just a few minutes, they have a great deal to do to earn a good ranking from Etsy’s algorithm and attract customers. First, they need to take flattering photographs of each of their products, preferably using a good camera, and sometimes they have to hire models or professional photographers who need to be paid. The ability to take pictures that correspond to the dominant aes- 126 thetic in the field is an effective way to gain a ranking as a reliable, efficient seller. Etsy sellers consequently spend a great deal of time poring over their competitors’ photographs, taking their own photographs, learning how to take good photographs, purchasing good camera equipment, and so on. Etsy sellers are also driven by ranking and visibility considerations to carefully tag each of their items in their online shop. Karine, a professional ceramist, spends a great deal of time trying to improve her Etsy shop’s online ranking: The Etsy system uses incredibly sophisticated algorithms. You need to use keywords. It took time to learn that. It’s not obvious to start with. The more effective your keywords, the higher up you appear in the search results. So, I gradually work my way up the rankings. It’s quite hard work. For example, for Mother’s Day, you have to add in the keyword ‘Mother’s Day’. With each change of season, the front image has to correspond to that season. And then there’s all the work on the Internet, which I love, all the work on screen, but it has to be done. To begin with, my products got completely lost. I had like two visits a month for a product. I couldn’t understand why. It was really when I finally took the time to read Etsy’s online training information that I understood where I was going wrong. It took me months to get there. (Karine, thirty-eight years old) Adjusting to Etsy’s algorithm takes time, and sellers need to gain a real grasp of the specific Etsy devices (such as keywords) if they want to be visible. Otherwise, just like Elisabeth, who sells textile products on Etsy, they ‘get a poor ranking’, are ‘completely invisible’, and ‘don’t make many sales’ (see part 6.1). Etsy provides its sellers with ‘shop stats’ consisting of a metrics overview (number of views, ‘favourites’, orders, revenue, etc.) and elements on how shoppers found the online shop. This scoreboard is used by Etsy sellers to check their visibility on the website. The Etsy website also asks its sellers to write narratives about both their products and themselves in order to personalize their online shop and make it more attractive to shoppers. Besides, Etsy sellers consider a project is finished not when they finish making an object, but when they have photographed the object and put it on their Etsy shop with a brief description. Here again, Etsy helps its sellers by providing them with a Seller Handbook with advice about the website and the associated smartphone application (‘beginner’s guides’, ‘finish your shop set-up’, ‘learn how to craft a great About section’, ‘optimizing your shop for Etsy searches’, and so on). Learning about online commodification also takes the form of peer discussions and forums. Begin- 127 ners can ask questions in the Etsy forums and learn how to manage online interactions with customers, how to set prices, how to read a business contract, and even how to respond in the event of a dispute. So although opening an Etsy shop does not appear to represent much of an investment at first, it takes skill, time, and resources to develop the shop. Online activity does not only concern the Etsy website and its smartphone application. The e-reputations of Etsy sellers depend on the frequency and regularity of their online publication of new items and posts as much on Etsy as on all the associated social media. Facebook, Instagram, Pinterest, and Twitter are key platforms these days on which to build a shop or a brand reputation. Etsy sellers have to learn how to use these social media strate- gically. Many of them also have a personal blog and/or a personal website (sometimes with e-commerce solutions such as Shopify), which they need to keep up to date. Posts and photos have to be replenished and renewed to drive traffic and attract potential customers to their Etsy webpage. Tue-Minh, a paediatrician who sells her paper sculptures as a hobby on Etsy, posts on Facebook and Instagram approximately every four days. She gradually learnt how to put together an effective post from her discussions with other creators: Over the year, as I associate with creators and the like, I’ve come to understand there’s another way to communicate about my creations. And I think I’m really influenced by the way others communicate too. I’ve come to understand how they do it, and so I copy them a bit. AJ: And what exactly is this way of communicating? I’ve realized it can be things that don’t have any direct association with a product. It’s better if the product’s there somewhere, directly or indirectly, but just to post about a product, like ‘the product is online’, does nothing. However, you have to have style. For example, once on my travels, I saw a concrete spiral staircase. It immediately looked to me like origami. And I thought, ‘That’s concrete origami’. I stopped to take a photo and then posted it as concrete origami. There’s a link there with what I do, i.e. a link with paper and all. So, it wasn’t to say, ‘Come and buy my stuff’. It was just to say, ‘This is my sensibility’, and therefore the artist’s behind it. It’s as if, with those photos, with social media, we were sharing a little bit of our sensibility for everyday life. (Tue-Minh, forty years old) An effective post is, therefore, a regular post, not only in terms of the products for sale, but also in terms of the lifestyle and artistic sensibility of the Etsy sellers who try to build an image of themselves as artists, creators or designers. 128 Narratives often talk about passion, authenticity, and creativity. Etsy sellers also carefully consider whether to publish pictures of themselves, disclose their job or town of residence, or talk about children and private family events (weddings, divorce, illness, etc.). Communication is therefore not just for the shops, but also concerns the Etsy sellers themselves. The value of home- made output is closely associated with the person behind the shop. The Etsy seller’s online persona is a controlled, selective, and sometimes glamorized version of their ‘true’ self. Specific self-presentation work (Goffman 1959) is paradoxically required to produce this online marketable ‘authentic’ self. 6.3.2 An Intensive, But Exciting Activity Online activity takes up a significant amount of time in an Etsy seller’s schedule, sometimes surpassing the time spent on the craft activity. Even hobbyists spend a few hours a week managing their online shop. Tue-Minh, the paediatrician who makes paper sculptures, considers she spends ‘easily two to three hours a day’ managing her Etsy shop and her Facebook and In- stagram accounts. She regrets that she does not have as much time to make her sculptures: ‘I’d like to spend more time crafting and less communicating. Right now, I communicate more than I craft.’ She also explains that this online time is hard to calculate: Actually, I don’t really know how much time it takes because it’s always on my mind. It’s like I’m always on the lookout for what I can post and whether I have something to say on social media. The act itself is very brief, as it only takes twenty minutes to post a picture with a caption. But the time in the making, as in what picture should I choose and how should I communicate, that’s the underlying side. It takes me time to think about it, and sometimes it doesn’t happen at all. (Tue-Minh, forty years old) Etsy sellers need a certain frame of mind to keep on the ball all the time, like a form of mental workload. This makes Etsy more than a mere leisure activity for hobbyists such as Tue-Minh: it represents ‘serious leisure’ (Stebbins 2007). Professionals spend more time than amateurs on online activity, and some of them are totally engrossed in it. Celine, who successfully sells silk-screen tee-shirts on Etsy, has incorporated her online activity into her everyday life as if it were ‘natural’: I’m always on my brand’s Instagram account, liking photos. I do it in front of TV series. I do it naturally, all the time, it’s quite natural. (Celine, twenty-three years old) 129 Professional time and domestic time are rolled into one with the ever-present smartphone, but Celine does not see it as a constraint. Aged twenty-three, she has developed a keen interest in the Internet and social media since childhood: I’ve always followed all that. After all, I am a Generation Yer, as I think they say. I’ve always loved fashion and style ever since I was little. I’ve always kicked about on the Internet, watching videos, blogs, websites, branding strategies. I find it all interesting. That’s why what I do now is pretty natural. (Celine, twenty-three years old) Many successful professionals on Etsy are young sellers like Celine, who master the codes of communication on the Internet. The generational aspect of these online skills also explains why older sellers sometimes look to their children or younger people than themselves to do or learn how to do certain tasks. Successful Etsy professionals dedicate themselves full time to their online activity, to the extent that they generally do not produce the items they sell. This is especially paradoxical considering that Etsy’s reputation is based on its promotion of DIY and handicrafts. The Etsy web scraping returned a finding that only 47 percent of the 1 percent happy few earning at least the minimum wage answer ‘I do’ to the question, ‘Who makes your products?’ (53 % answer ‘someone else’) compared with 85 percent for all Etsy sellers. This naturally concerns sellers of vintage items, but also sellers who outsource their production.2 During the interview, Celine made no secret of the fact that she had set up a partnership with a silk-screen printing workshop. Given that they do not produce themselves, Etsy sellers like Celine can focus on online communication and developing their ‘brand’, which is key to success. Paradoxically, success on Etsy calls for a shift away from the craft ethics pro- moted by the platform towards the intensification of online activity. The intensity of online activity is not only a question of the amount of time spent online. It also has to do with the emotional investment required. Given the importance of the pictures and narratives boasting ‘authenticity’ and ‘creativity’ (see part 6.2.1), Etsy sellers’ online activity can be equated with what has been conceptualized as ‘emotional labour’. According to A. Hochschild, who coined this concept, ‘This kind of labor calls for a coordina- tion of mind and feeling, and it sometimes draws on a source of self that we honor as deep and integral to our individuality’ (Hochschild 1983, 20). It is not easy for every Etsy seller to make the emotional investment and stage their feelings. Doing so calls for real skill and therefore, a learning process. Tue- 2 Vintage items can be sold alongside handmade products on Etsy. In 2013, Etsy changed its house rules to allow the outsourcing of production. This decision caused an outcry among Etsy sellers and buyers. 130 Minh explains that she used to be uncomfortable talking about her profession as a paediatrician in her online communication about her paper sculptures, even though her creator friends advised her to do so to personalize her shop: Lately, I’ve been selling more. I think it might be because I’m more visible now because I’m hiding less. I think that, subconsciously, part of me wanted to hide the fact that I was a paediatrician. Now, I’m more present on social media, I post more often, and I think I’m putting myself out there a bit more, showcasing a kind of lifestyle. (Tue-Minh, forty years old) This showcasing of the self is typical of emotional labour. And because it draws on emotions, or ‘sensibility’ as Tue-Minh puts it, this online activity is based on invisible gendered and domestic skills. This is why, in keeping with feminist tenets on the invisibility of mostly feminine ‘domestic work’, this labour has been conceptualized by post-feminist media studies as ‘women’s work’ (Jarrett 2015) regardless of the worker’s gender. The original feminine aspect of this kind of emotional labour is illustrated by the case of a married couple on Etsy. Jonathan defines himself as an artist making concrete sculptures, while his wife Audrey manages his Etsy shop and oversees communication. The interview was conducted with both of them and addressed the issue of the division of labour within the couple: AJ: In the media articles about you, Jonathan, do you appear alone or do you both appear? Audrey: Alone. Me, it’s like I don’t exist. I manage the communication, in inverted commas, the social media, and all that. So, I make out like I’m him all the time. Then I report back to him. I know how to pretend. He’s too real, too much himself. On social media, he hasn’t picked it up, because you have to be there pick up the codes and all. (Audrey and Jonathan, forty-one years old) Jonathan was less talkative than his wife during the interview. Audrey took charge of the interview, just like she manages the professional relations as- sociated with her husband’s shop. The intensive work she takes on is totally invisible (‘It’s like I don’t exist’), which is characteristic of ‘women’s work’. During my study, I met male sellers who were committed to this online ac- tivity to build their e-reputation, but I did not meet any male partners who took on this invisible work for their wife’s shop. Women are better conjugal resources than men when it comes to developing entrepreneurial activities. 131 The intensity of online activity is exhausting, but Etsy sellers also find it pleasurable and exciting. Celine, who sells silk-screen tee-shirts on Etsy, claims that she ‘loves it’. Hobbyists also find it exciting, as they associate online activity for their Etsy shop with an entrepreneurial game. Tue-Minh, who makes paper sculptures, enjoys ‘playing shopkeeper’ outside of her job as a paediatrician. Lucrative online activities tend to blur the conceptual distinction between ‘playing’, ‘gaming’, ‘working’, and ‘labouring’ (Lund 2014). This is why it is sometimes referred to as ‘playbour’ (Scholz 2012). 6.3.3 Is it Really Labour? Free Labour? The online activity described above can be defined as ‘work’ or ‘labour’. In an earlier co-edited book (Naulin and Jourdain, 2019), S. Naulin and I characterized online activity as an important part of the ‘extra work’ of commodifying a leisure activity, which extends beyond the mere activity of producing something for oneself, family, and friends. I would like to deepen this analysis here. At the risk of stating the obvious, the notion of ‘work’ or ‘labour’ is polysemous. M. Lallement (2001) distinguishes four main catego- ries: work-praxis associated with empowerment and achievement; work-status associated with social integration; work-suffering associated with efforts and punishment; and work-exploitation associated with the creation and extortion of value as in salaried employment. As an intensive and exhausting, but also an exciting activity fostering sociability, Etsy sellers’ online activity easily fits in with the three first conceptions of work. I would like to show that it is also a form of ‘work-exploitation’, and even a form of ‘free labour’, for professionals as much as for hobbyists. The association of Etsy sellers’ online activity with ‘digital labour’ and hence ‘free labour’ (Cardon and Casilli 2015; Scholz 2012) is not obvious. Like ‘work’, the concept of ‘digital labour’ developed in recent years refers to very different social practices. A large part of the literature on this concept focuses on private individuals who ‘like’ photos on Facebook or fill in a re- CAPTCHA for Google, and who, in doing so, add value to the companies behind Facebook and Google. These millions of individuals are therefore said to work for free (in the ‘work-exploitation’ sense) since they do not earn any monetary compensation from it. Hence the association of ‘digital labour’ with ‘free labour’ (Terranova 2000). Contrary to this description, Etsy sell- ers do not go online for free. They engage in online activity to commodify their products and to earn money. Does this mean that they develop a form of digital labour that is not free labour? I argue that their online activity is actually another form of free labour. 132 There is a discrepancy between the economic value generated by Etsy sellers’ activities and the income they earn in general. The economic value generated by their substantial online activity is unevenly shared among Etsy sellers, Etsy itself, and other companies behind the digital platforms. In part 6.1, I discussed the small amounts of money generally earned by Etsy sellers and the wide inequalities among them (average annual turnover of € 1,098 with a median of € 114). In 2015, the company behind the Etsy platform was listed on the stock exchange. Its turnover has grown steadily ever since, with the company posting $ 818 million in 2019. This turnover is made up of the 5 percent fees on each transaction and the $ 0.20 paid by sellers for each item they display in their online shop (if the item is not sold within four months, they have to pay again). Etsy is also increasingly monetizing rank- ings and e-reputation: sellers can buy options to improve their listing on the platform or use pay-per-click advertising on other platforms such as Google or Facebook. Today, such commodification of extra services accounts for a large part of Etsy’s revenue. As visibility on the platform is a scarce resource, sellers can either work on it or pay for it. When Etsy sellers communicate about their shop or brand on other web platforms such as Facebook and Instagram, they also promote Etsy. An advertisement for their online shop is advertising for Etsy. It may attract not only buyers but also new creators drawn to Etsy who want to join the ‘Etsy community’ by opening a shop and paying to display their first items. Etsy has developed a raft of devices such as the Seller Handbook to help Etsy sellers promote their shop and, therein, Etsy. Etsy is thus a co-creation and the economic value generated by Etsy sellers largely benefits the company behind the platform. In addition to the commodification of personal data, which also calls for consideration, the traffic generated by Etsy sellers’ posts on social media also benefits platforms such as Facebook, Instagram and Pinterest. Consequently, the economic value created by Etsy sellers is partially extorted from them, exactly as in the ‘work-exploitation’ concept, and part of their labour is unpaid. Their digital labour is, therefore, clearly a form of free labour. The fact that workers are dependent on the platform equates Etsy sellers’ activities all the more with ‘work-exploitation’. Although Etsy does not con- trol its sellers’ activities, as Uber does its drivers (Scholz 2016), many craft and design businesses depend on Etsy. Professionals generally have other sales channels than Etsy through which to earn a living from their output (fairs, physical shops, personal websites, and so on). Still, if Etsy were to close, their turnover would drop. This has already happened in France to the A Little Market handicrafts platform. Etsy bought the platform in 2014 and decided to close it in 2017 to prompt the platform’s sellers to make the 133 move to Etsy. The decision was opposed by A Little Market’s sellers who signed petitions, but nothing could be done to stop it. Many craft businesses had to close because the online activity they had developed was based on A Little Market’s specific metrics, which differed from Etsy’s. Another demon- stration of dependence on Etsy came in 2018 when Etsy decided to change its algorithm in order to foster local and national sales: French sellers with buyers in the United States and other foreign countries saw their sales drop and had to adjust their marketing strategy to fit in with the new algorithm. In general, Etsy sellers must constantly adapt to the platform’s new operating standards which correspond to a form an ‘algorithmic management’ (Wood et al. 2019). So, although Etsy is a co-creation, decisions are made solely by the company in the interests of its shareholders. Lack of recognition of the Etsy sellers’ contribution is another reason for terming this contribution ‘free labour’. As M. Simonet (2018) explains, free labour is labour that is both unpaid and socially unrecognized. Why do Etsy sellers consent to free labour? In fact, although their online activity is ‘free’, it is also a pleasurable activity that they enjoy doing. In keeping with T. Terranova, ‘free labour’ has two meanings: it is ‘not compen- sated by great financial rewards (it [is] therefore “free”, unpaid), but it [is] also willingly conceded in exchange for the pleasures of communication and exchange (it [is] therefore “free”, pleasurable, not imposed)’ (Terranova 2000, 48). Etsy sellers enjoy online activity as an exciting activity that brings them social benefits, such as visibility (via the shop’s exposure) and sociability (by means of forums and discussions among Etsy creators) (Naulin and Jourdain 2019). The two aspects of free labour—unpaid and pleasurable—go together and need to be analysed as two sides of the same coin. 6.4 Second Form of Free Labour: Traditional Domestic Work The second form of free labour does not apply to all Etsy sellers. It concerns the many women won over by the ‘new domesticity’ trend, who extol being at home. Most of them are hobbyists with a salaried job unrelated to crafts, but many are women who have left a salaried job to open a professional craft business on Etsy. These professional sellers see Etsy as an opportunity for a fulfilling work-life balance. They epitomize the ‘having it all’ ideal of post-Fordist societies that paints female micro-entrepreneurial homeworking in glowing colours (Adkins and Dever 2016). I probe this ideal and its actual impacts on Etsy sellers. 134 6.4.1 The ‘Having It All’ Ideal: Promise and Projects Etsy’s promise of women’s empowerment works all the more as it ties in with post-Fordist values such as, ‘domesticity, familism, entrepreneurship, bound- less love, heteronormative femininity and intimacy, excessive attachments to work, indebted citizenship and financial literacy’ (Adkins and Dever 2016, 3). Sometimes, the very creation of an Etsy shop is itself driven by the aspi- ration to become an entrepreneur with one’s own business. Beate, who left her job as a consultant in a large corporation and now describes herself as a ‘mumpreneur’ with two daughters, is particularly clear about her initiative to start a crocheted baby clothes business on Etsy: Crochet was not really […] To tell you the truth, I know I write in my shop that it’s always been my passion, but that’s not true. That was driven by the need to find an idea. I wanted to launch my own business. (Beate, forty-six years old) Like Beate, some Etsy sellers without craft skills have taught themselves to make jewellery and bags, mainly from web tutorials, just to be able to open a shop on Etsy. The figures of maker and creative entrepreneur have become new ideals for women (Taylor and Luckman 2018). Occupational shifts are frequent among female Etsy sellers. Launching a craft business is a way to break away from unsatisfactory working conditions. Such is the case with Elisabeth, a mother of three, who left her job as an IT engineer to make and sell textiles for children (bags, blankets, etc.). During the interview, she was particularly critical of her working conditions in a large corporation. Her three-year break on parental leave made those working conditions even harder to bear: I didn’t like it anymore. I switched off. I didn’t feel comfortable anymore. And it was complicated because there wasn’t much of a future there. I always did the same work. Every higher position was already taken, so I didn’t have many prospects. And it was complicated being a girl, as I often got put in charge of documentation ‘because girls like documentation’. (Elisabeth, forty-two years old) Beate and Elisabeth’s cases are representative of many women’s occupational shifts to craft jobs. The individuals concerned are mainly educated women who work for large companies as consultants, human resources managers, engineers, sales representatives, and so on. Even though they earn a relatively comfortable income, these women can be highly critical of their occupation: they talk about a loss of interest in their work, mainly because they have no 135 command of the entire production process and sometimes because they cannot bear the chain of command in their firm (Jourdain 2014). They often feel frustrated with their mundane jobs, especially when they have studied for years to get there. Their frustration is also due to the lack of career oppor- tunities in their workplace. This is a gendered situation that is due partly to the glass ceiling and partly to the general deterioration in working conditions and terms of employment in contemporary societies (Méda and Vendramin 2017). Starting a craft business is a way to escape this frustration, to do a meaningful, rewarding job, and to fulfil self-employment aspirations. Maternity leave and parental leave are frequent hiatuses during which women develop the marketization of a leisure activity (Naulin and Jourdain 2019). Hiatuses are characterized by a lack of formal employment and forms of social isolation. They represent ‘turning points’ (Hughes 1971) in women’s trajectories. Seven of the sixteen women in my qualitative sample of Etsy sellers started their leisure activity while on maternity or parental leave when they had the free time to develop a new activity at home. Like Beate and Elisabeth, their new activity is sometimes associated with motherhood, such as making baby clothes. Many women learnt handicrafts in the domestic sphere in childhood and adolescence from a relative, mother, or grandmother. Others learnt their skills ‘on their own’, from scratch, using web resources (blogs, forums, tutorials, etc.) and, in some cases, by attending workshops and crafts courses. During the hiatus, these skills are (re)used to form both a leisure and a domestic activity. Marketization on Etsy is then a means to make domestic activities productive, streamlined and potentially profitable. It can be seen as making reproductive work pay. After the maternity or parental leave, the commodified activity is sometimes raised to the level of a professional undertaking. Women who leave their job to become stay-at-home working mothers have been dubbed ‘mumpreneurs’, a neologism coined from ‘mum’ and ‘entrepreneur’. C. Ekinsmyth defines them as, ‘female entrepreneurs who operate at the interface between paid work and motherhood’ (2011, 104). Even though she does not define herself as a ‘mumpreneur’ (as Beate does), Elisabeth has all the characteristics of one. She decided to leave her job to escape unsatisfactory working conditions, but also to take better care of her three children: I said to myself, ‘I’m going to do it differently’. I started sewing on parental leave, so I thought, ‘That’s what I’ll do’. I was a bit naïve. I didn’t understand the implications of it all, ‘It’s going be great! I’m going to work from home. That way, I’ll have time for my children, 136 and I’ll be able to work and sell and make money.’ (Elisabeth, forty- two years old) This excerpt is a perfect example of the ‘having it all’ myth, which is particularly widespread among Etsy sellers. Elisabeth only works at her business during school hours, from 9am to 4:30pm (except Wednesdays and weekends). She is happy to spend a lot of time with her children, especially since their father gets home late at night. However, as I will show in part 6.3.3, she does not consider her activity to be a form of empowerment anymore. Some female Etsy sellers choose to hold onto a salaried job alongside their activity on Etsy, defining themselves as hobbyists. Many are also won over by the narratives about new domesticity, creativity and entrepreneurship. They sometimes regard commodification on Etsy as a way to offset discontent at work (boredom, adverse working conditions, frustration, stress, etc.). Diane, who works as a marketing manager for a start-up, created her shop on Etsy (selling textile jewellery) because she was bored with her previous job in a large company: I worked for a British laboratory, which was taken over by an American lab. And I said to myself, ‘A crushing office life is not for me’. And something was missing. I have desires, I felt I could do more enjoyable things. I didn’t feel fulfilled. I loved sewing. And then, I thought, ‘Take the plunge and create your own online shop’. (Diane, twenty-nine years old) The belief in self-fulfilment through creative business is pervasive among Etsy sellers, both professionals and hobbyists. However, they are not naïve and many of them know they will not earn a living on Etsy. Those who do not have a partner who can finance their unprofitable activity prefer to keep a salaried job, even one that is unsatisfactory and to find self-fulfilment in their leisure time. 6.4.2 In the Name of the Family: Gender Inequalities Within the Household Domesticity is necessarily associated with traditional domestic work. Wom- en who decide to open a craft business on Etsy to work from home have to subscribe to taking on domestic responsibilities, such as housekeeping and childcare. In the case of those making an occupational shift from a salaried job to self-employment at home, such tasks used to be performed by other people: sometimes a housekeeper and more generally babysitters and the childcare system. Male partners also used to be more present to help out with 137 domestic duties, albeit without any gender equality. They would sometimes take the children to school in the morning, for example. Women who make this career change are expected to take on all these responsibilities. House- keeping and childcare costs can moreover play a role in persuading some couples that a stay-at-home business is a good idea. Traditional domestic labour is free labour that benefits male partners. Whereas digital activity has only quite recently been equated with free la- bour, feminist gender studies have been presenting domestic labour as free work for at least forty years (Jarrett 2015). The feminists’ tour de force was to label as ‘work’ that which was not previously understood as work, i.e. women’s domestic activities. Forty years later, this is still inspiring analyses of ‘new domesticity’. The value created by this reproductive work is captured essentially by male partners, as it is in traditional Fordist societies based on the ‘breadwinner’ family model. Elisabeth explains that her career change benefits her husband, who encouraged her to make the shift: To a certain extent, he [her husband] is quite happy because all he does is run his own life. He leaves when he wants, he comes back when he wants. I’m here anyway. Now when the children are sick or when they have an appointment, I’m the one who takes care of everything. (Elisabeth, forty-two years old) This management of ‘everything’ in the household is unpaid and benefits the husband who can focus on his own executive career. As free labour, it is also ‘invisible work’ (Kaplan Daniels 1987), all the more so since it is obscured by the handicrafts and marketization activities. This family set-up also makes the women more financially dependent on their partners, which can be problematic in the event of a divorce (Jourdain 2014). In the interviews with female Etsy sellers, the drudgery of domestic tasks is put into perspective in terms of the ‘quality of life’, especially with respect to their children. Most women explain their career change based on their desire to spend more time with their children. They enjoy picking up their children from school at 4:30pm and being in charge of their education. In the name of the family, they take on and even take pleasure in invisible domestic work, which is typical of free labour (Simonet 2018). 6.4.3 Failing It All? Despite Etsy’s promise and female Etsy sellers’ intentions, no one manages to ‘have it all’, that is to run a successful authentic craft business and cope with every domestic facet in the household. Mumpreneurs who have to work 138 during their children’s school hours cannot simultaneously keep house, craft their items, and perform an intensive online activity to sell on Etsy. Many of them are disappointed, like Elisabeth, who had hoped she could earn more money from her textiles (see part 6.1.1). She now considers that she cannot ‘have it all’, but she does appreciate the time she spends with her three children: I make less than I spend. I’m lucky to have a husband who can provide for us, so I can afford to do it. Because my goal is obviously to be profitable, but also to be able to take care of my children while keeping a foot in the workforce. I’m well aware I can’t do both. I can’t do both to the best of my ability. So, I take care of the kids the best I can, and I do the best I can for my shop. (Elisabeth, forty-two years old) Elisabeth had previously described herself as ‘naïve’ about her career change (see part 6.3.1). She now realises that her project was unrealistic and that she cannot ‘do both’, that is have a successful craft business and be an active housewife and mother. An examination of successful Etsy sellers’ profiles proves that the ‘having it all’ ideal is nothing but a myth. The happy few, who earn more than the minimum wage on Etsy (1 % of all sellers), are more often than not men, and those I met had no children. Indeed, the intensive online activity demanded to commodify on Etsy, as described in part 6.2, is not compatible with childcare and housekeeping. This is why, unlike the media image, mumpreneurs are not successful Etsy sellers. I actually met a quite successful Etsy seller with two children, but she was far from being an example of the mumpreneur philos- ophy since the domestic work was handled by her husband. This household organization was presented by Sandra as a necessary condition to engage in a craft business on Etsy: God knows he [her husband] made sacrifices for that because our family life changed completely. On weekends, I would get up at 7am, work all day and go to bed at midnight. So, no more family life, no more going out, no more nothing. We cut ourselves off from a lot of people who didn’t understand. My husband, he’s the one who cooks, who does the shopping, who cleans. So, our life was turned completely upside down. And it’s been difficult for our children too. They are twelve and sixteen years old now, but my son was 7 when I started. And the number of times I heard him say, ‘Anyway, I know you prefer sewing to your children’. That’s really hard to hear. (Sandra, forty-two years old) 139 This excerpt is filled with regrets. Sandra’s couple is atypical of the traditional distribution of tasks, and she considers that time with her children has been sacrificed to her craft business. This inverse profile shows that women cannot balance a successful craft business with an active family life. Their entourage encourages them to prioritize the latter, as shown by the general disapproval of Sandra’s way of life. Where balancing business and family life is already a tall order, balancing crafts, business, and family life becomes herculean. I have already shown that the majority of successful Etsy sellers do not craft the items they sell them- selves: they outsource production to spend their time on the online activity (see part 6.2). Mumpreneurs like Elisabeth, who sews her own textiles, cannot compete with these sellers. Domestic work and crafting prevent them from working on their brand’s reputation, resulting in poor rankings and poor sales on Etsy. Although they may identify with the media image of the successful Etsy seller—or rather its glamorized version—they do not actually have the same profile or the same activities. Domestic work takes time and forms an underlying factor in the lack of recognition of the business activity. Family and friends often view making and selling items on Etsy as a mere leisure activity, even as mumpreneurs struggle to get it recognized as real work. They do not like being taken for stay-at-home housewives. Elisabeth explains that this misconception com- plicates her business: The big problem when you work from home is that other people don’t understand that in the phrase ‘work from home’, you find the word ‘work’. When you’re at home, people think you’re available, and it’s very testing because your friends have their own schedules. They don’t necessarily understand that I may well be at home, but I’m working. And they call me, ‘Are you coming over for a coffee?’ ‘No, I’m not’. It’s a really awkward issue. If I go, it’ll set me back a day. That’s the effect I didn’t fully anticipate. It’s really awkward with some friends who think that when you’re at home, you don’t do anything. (Elisabeth, forty-two years old) The view of a craft business as not being real work can even be found among close friends and husbands. And this unexpected lack of recognition is at odds with the supposed empowering effect of the marketization of a leisure activity. Why do these women persevere regardless? Despite the many difficulties, they appreciate their social benefits. Most importantly, they value the time spent with their children and the ‘quality of life’ made possible by their mi- croenterprise, but also by their partner’s regular income. They generally make 140 new friends with other creators and mumpreneurs on online forums and in meetings with local Etsy teams, many of which are run by mumpreneurs. These new likeminded friends understand their way of life and the difficulties they face more than their old friends who work as salaried employees. This new sociability is valuable and encourages them to keep going with Etsy. Finally, these women cannot ‘have it all’: they have to choose between family, crafts, and business. Any attempt at ‘having it all’ could end up with them ‘failing it all’. 6.5 Conclusion: The Development of (Free) Work for Marketplace Platforms in Pandemic Times The case of Etsy is symptomatic of the widespread transformations in work in capitalist societies driven by the development of the digital economy. Indeed, Etsy is typical of many other marketplace platforms, such as Airbnb, Amazon Handmade, Craiglist, eBay, TaskRabbit, Upwork, Vinted, but also many other less known e-commerce websites. The supply side of these platforms is made of a multitude of private individuals who are eager to earn ‘extra money’ thanks to the online commodification of their products or services (Naulin and Jourdain 2019). Some of them even hope to earn a living or at least substantial gains to cope with precarity when they register to one of these websites. The degree of professionalization varies from platform to platform: it is higher for Upwork and lower for eBay or Vinted. Even though Etsy concerns a majority of hobbyists, it also includes many (aspiring) profes- sionals. Therefore, this case study enables us to explore the variety of sellers’ profiles in today’s digital economy. Etsy also has the same characteristics of many other platforms: ‘algorithmic control’ (Wood et al. 2019), marketing of the ‘entrepreneurial ethos’ (Ravenelle 2017), limited ‘firm’s obligation to the workforce on which it relies’ (Vallas and Schor 2020), and so on. Still, Etsy has its specificities: it is one of the most feminized platforms and, unlike many platforms (except Upwork), it does not sell standardized products or services but rather ‘singularities’ (Karpik 2010), which compels its sellers to constantly showcase the quality of their objects. In this sense, Etsy sellers are close to another type of platform workers: ‘content producers and influencers’ (Vallas and Schor 2020), like bloggers or vloggers who commit themselves to ‘aspirational work’ (Duffy 2017). The characteristics of the work analysed through Etsy’s case study are thus representative of a great number of platform workers, and especially the women coming from the middle-class. 141 In line with previous studies on digital female workers who are ‘not getting paid to do what [they] love’ (Duffy 2017), my research highlights a paradox: although platform workers set out to earn money from commodification on a platform, they have to engage in extra work, which is unpaid and not socially recognized, and therefore constitutes free labour. Online activity, of benefit to the platform companies, and domestic work, of benefit to male partners, are forms of free labour. They are perverse effects of marketization on the marketplace platforms. This free labour makes platform workers, especially women, even more insecure and dependent, not to mention the risks they incur and their lack of social protection (Ravenelle 2019). Consequently, the digitization of work, associated with the neoliberal valorization of the ‘enterprise of the self ’, reinforces the precarious situation of women in con- temporary societies, both at work and at home. How has the Covid-19 pandemic affected marketplace platforms and thus their workers? The answer might differ from one platform to the other, according to the specificities of the products or the services sold. Overall, the pandemic seems to have accelerated existing trends, such as the development of the gig economy and the associated insecurity. Since companies support- ing marketplace platforms ‘have prioritised profit and customers’ needs over workers’ safety’ (Alvarez de la Vega et al. 2020, 1), working conditions have deteriorated. More research is required to present robust results on the eco- nomic and social effects of the pandemic. My field study on Etsy, which was supplemented by new data in 2020 and 2021, can be a contribution to this opening research. In France, the number of Etsy sellers increased by 74 percent in 2020, and this figure is also very high in other countries, especially in the United-States. While many sellers experimented difficulties with their online shop during the first lockdown from March to May 2020 in France—either because they could not easily send their packages or because they had to take care of their young children while schools were closed—lockdowns, curfews, and political incentives to stay at home contributed to the increase in the number of online shops. On 3 April, Etsy sent an incentive to its sellers all over the world to make facemasks. The facemasks trade contributed to the sales take-off on the platform, but all categories of items have developed. In 2020, Etsy’s Gross Merchandise Value (GMV) increased by nearly 90 per- cent compared to 2019, while the increase was only 20 percent per year in previous years. Etsy’s stock price has also skyrocketed since 2020. As a result, the company really benefited from the pandemic. On the workers’ side, the craze for opening an online shop can be explained by the extraordinary con- text of the Covid-19 crisis. The pandemic entailed economic difficulties for many individuals who lost their jobs and, at the same time, increased home 142 leisure time. For some, opening an Etsy shop might have been a way to earn extra money in a time of economic crisis. Moreover, as Sidonie Naulin and I showed in a previous text, the online commodification of domestic and leisure activities often begins during a hiatus ‘characterized by a lack of formal employment and/or forms of social isolation’ (Naulin and Jourdain 2019, 65). Whereas hiatuses used to concern women on maternity, parental or sick leave, jobseekers, expatriates, pensioners, and students, the pandemic has extended the hiatus experience to everybody but those working in ‘essential’ sectors (like health, care, or food). Many men and especially women chose to value their extra domestic time by taking up handicrafts and opening an Etsy shop. With the pandemic, many individuals also expressed the desire for a career change, including in the craft business as it is promoted by Etsy. Therefore, more and more free digital labour has been provided to Etsy, which contributed to the company’s expansion. Now, the following question arises: will these new platform workers continue to work for the digital economy? This question is linked to another one, outside of the pandemic context: why would workers consent to the free labour required by marketplace platforms and persevere with their online business? Actually, most of them do not, as I have shown in the case of Etsy, where turnover was already very high between 2017 and 2019. Many pandemic-time sellers may also stop their online activity when they realise that they cannot make a lot of money on Etsy. Amateurs, in particular, who are less dependent on the platform, are extremely mobile. The other workers enjoy certain aspects of the free labour, which is also seen as a pleasurable activity (fun and exciting), and consent to it in the name of higher values (reputation and family) typical of post-Fordist societies. They also extol the social benefits associated with their commodification activity (visibility, sociability and quality of life). Lastly, professional workers, especially those who leave a salaried job, champion the many advantages of self-employment and tend to gloss over the abovemen- tioned disadvantages. This positive view of self-employment, which might evolve with the pandemic, can be analysed as an indicator of the general deterioration in working conditions in salaried employment. 6.6 References Abdelnour, Sarah and Dominique Méda (Eds.). 2019. Les Nouveaux travailleurs des applis. Paris: Presses Universitaires de France. Adkins, Lisa and Maryanne Dever (Eds.). 2016. The Post-Fordist Sexual Contract: Working and Living in Contingency. Basingstoke: Palgrave Macmillan. 143 Alvarez de la Vega, Juan Carlos, Marta E. Cecchinato, and John Rooksby. 2020. The Gig Economy in Times of COVID-19: Designing for Gig Workers’ Needs. Microsoft Research Symposium on The New Future of Work, Virtual Symposium. August 05‒07, 2020, https://www.microsoft.com/en-us/research/uploads/ prod/2020/07/NFW-Alvarez-de-la-Vega-et-al.pdf (accessed March 3, 2021). Anderson, Chris. 2006. The Long Tail: How the Future of Business is Selling Less of More. New York: Hyperion Books. Beuscart, Jean-Samuel, Pierre-Marie Chauvin, Anne Jourdain, and Sidonie Naulin. 2015. La réputation et ses dispositifs: Introduction. Terrains & travaux, 26(1): 5‒22. Botsman, Rachel and Roo Rogers. 2011. What’s Mine is Yours: How Collaborative Consumption is Changing the Way we Live. London: HarperCollins Business. Cardon, Dominique. 2015. À quoi rêvent les algorithmes : Nos vies à l’heure des big data. Paris : Seuil/République des idées. Cardon, Dominique and Antonio Casilli. 2015. Qu’est-ce que le Digital Labour ? Paris: INA. Casilli, Antonio. 2019. En Attendant les robots : Enquête sur le travail du clic. Paris: Seuil. Cochoy, Franck and Sophie Dubuisson-Quellier. 2013. The Sociology of Market Work. Economic sociology_the european electronic newsletter, 15(1): 4‒11. Delphy, Christine and Diana Leonard. 1992. Familiar Exploitation: A New Analysis of Marriage in Contemporary Western Societies. Oxford: Polity Press. Duffy, Brooke Erin. 2017. (Not) Getting Paid to Do What You Love: Gender, Social Media, and Aspirational Work. New Haven, CT: Yale University Press. Ekinsmyth, Carol. 2011. Challenging the Boundaries of Entrepreneurship: The Spatialities and Practices of UK ‘Mumpreneurs’. Geoforum, 42: 104‒114. Flichy, Patrice. 2017. Les Nouvelles Frontières du travail à l’ère numérique. Paris : Le Seuil. Foong, Eureka, Nicholas Vincent, Brent Hecht, and Elizabeth M. Gerber. 2018. Women (Still) Ask for Less: Gender Differences in Hourly Rate in an Online Labor Marketplace. Proceedings of the ACM on Human-Computer Interaction, 2(53): 1‒21. Goffman, Erving. 1959. The Presentation of Self in Everyday Life, New York: Anchor Books. Hochschild, Arlie R. 1983. The Managed Heart. Berkeley: University of California Press. Hughes, Everett. 1984 [1971]. The Sociological Eye: Selected Papers. New Bruns- wick: Transaction Books. Huws, Ursula. 2003. The Making of a Cybertariat: Virtual Work in a Real World. New York: New York University Press. 144 Jarrett, Kylie. 2015. Feminism, Labor and Digital Media: The Digital Housewife. New York: Routledge. Jourdain, Anne. 2014. Du Cœur à l’ouvrage : Les artisans d’art en France. Paris: Belin. Kaplan Daniels, Arlene. 1987. Invisible Work. Social Problems, 34(5): 403‒415. Karpik, Lucien. 2010. Valuing the Unique—The Economics of Singularities. Princ- eton: Princeton University Press. Lallement, Michel. 2001. Daedalus laborans. Revue du MAUSS, 18(2): 29‒49. Lund, Arwid. 2014. Playing, Gaming, Working and Labouring: Framing the Concepts and Relations. tripleC, 12(2): 735‒801. Matchar, Emily. 2013. Homeward Bound: Why Women Are Embracing the New Domesticity. New York: Simon & Schuster. Méda, Dominique and Patricia Vendramin. 2017. Reinventing Work in Europe: Value, Generations and Labour. Basingstoke: Palgrave Macmillan. Naulin, Sidonie and Anne Jourdain (Eds.). 2019. The Social Meaning of Extra Money: Capitalism and the Commodification of Domestic and Leisure Activities. Basingstoke: Palgrave Macmillan. Ravenelle, Alexandrea J. 2017. Sharing Economy Workers: Selling, Not Sharing. Cambridge Journal of Regions, Economy and Society, 10(2): 281‒295. Ravenelle, Alexandrea J. 2019. Hustle and Gig: Struggling and Surviving in the Sharing Economy. Berkeley, CA.: University of California Press. Scholz, Trebor. 2012. Digital Labor: The Internet as Playground and Factory. New York: Routledge. Scholz, Trebor. 2016. Uberworked and Underpaid: How Workers are Disrupting the Digital Economy. New York: Polity Press. Schor, Juliet. 2018. ‘The Platform Economy: Consequences for Labour, Inequality and the Environment.’ In Max Neufeind, Jacqueline O’Reilly, and Florian Ranft (Eds.), Work in the Digital Age: Challenges of the Fourth Industrial Revolution. London: Rowman & Littlefield International. Simonet, Maud. 2018. Travail gratuit : la nouvelle exploitation ? Paris: Textuel. Stebbins, Robert. 2007. Serious Leisure: A Perspective for Our Time. New Jersey: Transaction Publishers. Taylor, Stephanie and Susan Luckman (Eds.). 2018. The New Normal of Working Lives: Critical Studies in Contemporary Work and Employment. Basingstoke: Palgrave Macmillan. Terranova, Tiziana. 2000. Free Labor: Producing Culture for the Digital Economy. Social Text, 63(18‒2): 33‒58. Vallas, Steven P. and Juliet Schor. 2020. What Do Platforms Do? Understanding the Gig Economy. Annual Review of Sociology, 46(1): 273‒294. Wood, Alex J., Mark Graham, Vili Lehdonvirta, and Isis Hjorth. 2019. Good Gig, Bad Gig: Autonomy and Algorithmic Control in the Global Gig Economy. Work, Employment and Society, 33(1): 56‒75. 145 Zelizer, Viviana. 1994. The Social Meaning of Money: Pin Money, Paychecks, Poor Relief, and Other Currencies. Princeton: Princeton University Press. 146 7 Ambivalent Flexibility: Crowd Workers Between Self-Determined Work and Capitalist (Self-) Exploitation on Digital Platforms Eltje Gajewski 7.1 Introduction Crowd work is a new form of work organization in which companies, orga- nizations or individuals use digital platforms to place work orders of various kinds with unknown contractors who then execute the respective order in return for payment (Eurofound 2015). The consequences of this new form of gainful employment for working subjects are widely discussed. On the one hand, the novel scope for flexibility, compatibility, and autonomy is emphasized, while on the other hand, the working conditions and the consequences for the entire working world are problematized. How crowd workers themselves experience their activities on digital platforms, however, has hardly been re- searched. In addition, there is a lack of relevant theory-based and -generating empirical research that analyses the significance of crowd work in the context of a changing working society. The present chapter takes up these gaps and works out the ambivalent character of flexibility of digital platform work for the crowd workers against the background of their own demands. The focus will be on the novel scope that work on digital platforms certainly opens up, as well as on the mechanisms of delimitation and exploitation, and the self-rationalization with which the employees ultimately react to the latter. On this basis, an answer to the question of how crowd workers deal with the ambivalent character of flexibility on crowd work platforms is to be found. This chapter is based on the Master’s thesis ‘Ambivalente Flexibilität: Crowd worker*innen zwischen selbstbestimmter Arbeit und kapitalistischer (Selbst-)Verwertung auf digitalen Plattformen’ (Gajewski 2018), in the con- text of which four crowd workers on two German platforms were questioned 147 about their work, the working conditions they found on digital platforms, and their own demands using semi-structured interviews. These were then analysed with the help of the Documentary Method. The main finding is that the activity on digital platforms has an ambivalent character for the subjects working there: they have special demands for flexibility and autonomy in their gainful employment, which they anticipate as already realized in crowd work, and for this very reason react increasingly with self-rationalization and adaptations to the work requirements on digital platforms. The scope for action actually opened up by digital platform work thus turns into an instrument of more efficient labour exploitation—because it is internalized and applied to oneself. The chapter is divided into six sections. After this introduction, the second section clarifies the central concepts and introduces the subject area of crowd work. In this context, relevant findings regarding the labour law situation and the basic working conditions on german digital platforms are presented and the current state of research on the subjects of work on these platforms—the crowd workers—is reviewed. In the third section, the theoretical framework is presented, which serves as a perspective for the empirical investigation and the research question. The fourth section then describes the research design. Finally, in the fifth section, the empirical results are presented and analysed. In the conclusion, the central findings are recapitulated. 7.2 Definition and State of Research 7.2.1 Definition of Crowd Work Crowd work is a new form of work organization in which a wide variety of actors publish various work orders via special digital platforms for a mass of unknown contractors, who then accept the orders self-selectively and produce the respective work services in return for payment. Crowd work activities can be commissioned by companies, public institutions (Felstiner 2011), non-profit and charitable organizations, government agencies or private in- dividuals (Leimeister and Zogaj 2013). The range of job types extends from work with a low level of complexity and very short processing time to more complex and creative activities, e.g. the creation of advertising copy. Crowd work is explicitly outlined as a paid form of crowdsourcing (Felstiner 2011). Crowdsourcing is a superordinate category that encompasses all forms of outsourcing of activities by means of a call on special IT platforms to external, independent, and self-selecting actors, whereby these platforms are operat- ed either by the clients themselves or by an independent service company. 148 Crowd work is distinguished from unpaid forms where intrinsic/value-based motives are the decisive factor for participation. Crowd work should also be distinguished from gig work.1 While the work orders in gig work are arranged via digital platforms such as Uber, the actual service is provided offline, i.e. in the ‘physical world’. Crowd work, on the other hand, is characterized by the fact that the orders are both mediated on the digital platform and carried out entirely within the platform environment (Durward et al. 2016; Gerber and Krzywdzinski 2019; Kenney and Zysman 2019; Leimeister et al. 2016; Menz and Cárdenas Tomazic 2017). According to Durward et al. (2016), another key aspect of the definition is that the income from crowd work activities make an indispensable contribu- tion to the livelihood of the crowd workers—regardless of whether they work full-time or part-time—and that the clients do not pay any contributions to social insurance or pension funds (Durward et al. 2016). Finally, according to this definition, crowd workers are considered economically independent and can decide independently on their working hours and content (Durward et al. 2016; Leimeister et al. 2016). It is also elementary that the provision of services is handled completely or to a large extent via a corresponding, specialized specialized IT platform (Durward et al. 2016). For the definition of different types of jobs, a rough distinction is made between microtasks, which have such a low level of complexity that they can be completed within a few minutes by unskilled persons—for example, the labelling of images—and more complex or even creative activities—for example, text creation (Kuek et al. 2015). 7.2.2 Labour law and working conditions Crowd workers are considered self-employed in Germany for several rea- sons. The jobs are advertized by means of a call on an online platform; the crowd workers search the platforms for suitable jobs and then accept them on a self-selective basis. In doing so, they are relatively autonomous in the execution of the accepted assignment, as they can decide when and where they work on the completion of the job and also use their own tools—that is their private digital devices. Only a certain processing period or an end date is usually set (Däubler and Klebe 2015). Crowd workers, therefore, do not have the same standards as employees, that is protection under labour law, collective agreements, sick pay, social security insurance, holiday entitlements, and the possibility of parental leave; at the same time, however, they are more dependent on the contractors and 1 For an overview, see Stanford (2017). 149 platform operators than ‘genuine’ self-employed persons (Benner 2016; Menz and Cárdenas Tomazic 2017). In contrast to dependent employment, in crowd work, it is not the sale of labour that is paid for, but the finished product or work result, although there is also the possibility that the work results are rejected by the clients without a reason and payment is therefore refused. It is often criticized that the rights of use of all submitted proposals for solutions are usually transferred to the client—even those that have not been remunerated (Benner 2016; Menz and Cárdenas Tomazic 2017). Another integral part of the work on crowd work platforms are the repu- tation systems. With these tools, clients can rate the work of crowd workers who have completed jobs for them. The number of positive evaluations is crucial for crowd workers to accept further crowd work orders, as the plat- form mechanics often regulate access to orders via ranking systems and crowd workers with a higher rank are either shown more orders or are allowed to accept more orders within a certain period of time (Benner 2016; Pongratz and Bormann 2017). Contact with the clients is normally only possible via the platform mech- anism and is therefore highly anonymous and standardized (Benner 2016; Schörpf et al. 2017). The relationship of crowd workers to each other is, on the other hand, rather characterized by a competitive relationship and is further alienated by the anonymity of the internet (Benner 2016). Some platforms even explicitly prohibit contact with other crowd workers in their general terms and conditions (Benner 2016). On this basis, (organized) co-determination is hardly possible for crowd workers (Menz and Cárdenas Tomazic 2017; Pongratz and Bormann 2017). 7.2.3 Crowd Workers: Previous State of Knowledge In general, it can be stated that crowd workers in Germany are predomi- nantly young, single persons with a mostly high level of education (Al-Ani and Stumpp 2015; Bertschek et al. 2015; Leimeister et al. 2016). Irrespec- tive of the field of activity, three-quarters of those surveyed by Leimeister et al. (2016)2 have a high school diploma and almost half of all the crowd workers in question have a university degree, a degree from a university of 2 Two samples serve as a data basis: The total sample comprises 434 persons. From this sample a further sample with n = 248 was generated, from which all respondents working on the Mylittlejob platform were excluded. This platform is explicitly aimed at students, so that distortions in the overall sample (e.g. with regard to the educational level as well as the distribution of microtasks, as these are often used as student extra income due to easy access) can occur in this respect. The following comments are based on this latter sample (Leimeister et al. 2016). 150 applied sciences, or a bachelor’s degree, although trainees and students are more often employed in the microtask area (Leimeister et al., 2016). Most crowd workers are freelancers, full-time employees, students or self-employed; part-time employees, pupils, apprentices, housewives, and pensioners are less active as crowd workers (Al-Ani and Stumpp 2015; Bertschek et al. 2015; Leimeister et al. 2016). With regard to the total income of crowd workers, it can be stated that the majority have a household netto income of less than € 1,500 per month, although a wide range of different wage classes below the poverty line is un- fortunately not reflected in the available studies (Al-Ani and Stumpp 2015; Bertschek et al. 2015). About two-thirds of the surveyed crowd workers state that they use the gains from their crowd work activities as additional income (Leimeister et al. 2016). The weekly working time depends on whether the income is used as main or secondary income. On average, it amounts to ten hours per week, although the variance is also large here (Leimeister et al. 2016, 51). The majority of crowd workers (over 80 %) work from home (Leimeister et al. 2016, 54). If the crowd working activity is carried out on a full-time basis, there is often a lack of insurance cover. More than one-third of full-time crowd workers are not insured against illness and/or unemploy- ment and almost half are not insured for old-age provision (Leimeister et al. 2016, 56; 58). 7.3 Theoretical Approach In order to make crowd work and the concept of flexibility theoretically more tangible, the theoretical arguments of the ‘new spirit of capitalism’ (Boltanski and Chiapello 2003) will be used. This theory focuses on the initial question of how the continued existence of capitalism, even against resistance, can be explained. According to Boltanski and Chiapello, there is no intrinsic motivation for either wage-earners or capitalists to continue the capitalist production process described above: The former have no share in both the means of production and the products of their labour and are thus dependent on subordination to the ruling class, while the latter are forced to continue the process of accumulating surplus value in constant competition with one another. In order for capitalism to exist as a comprehensive context of action, the (co-)work of the actors living within it is necessary. By means of a his- torically specific, changing ideology—a consciousness objectively necessary for the functioning of societies, which has grown out of the social conditions themselves and leads to the subjects being able to evaluate social reality even 151 in the categories of the respective form of society—these are motivated to continually participate in the capitalist production process (Boltanski and Chiapello 2003; Boltanski and Chiapello 2001; Eichler 2009; Kocyba 2005). This ideology is called the ‘spirit of capitalism’.3 Boltanski and Chiapello, however, note the existence and effectiveness of several historically changeable ‘spirits’ of capitalism, which, depending on the contemporary form of capitalism, have different incentive structures, orders of justification, and modes of security (Boltanski and Chiapello 2001). For the current form since the 1980s and 1990s, they diagnose a new form of company work organization in which industrially or bureaucratically and hi- erarchically organized structures have been exchanged for network structures, project-oriented employment relationships, relocation and outsourcing, flat hierarchies and extended scope for autonomy, and self-realization potential (Boltanski and Chiapello 2001; Deutschmann 2008; Nies and Sauer 2012). According to the authors, this change took place on the basis of capitalism’s appropriation of so-called artistic criticism. Boltanski and Chiapello distin- guish two relevant forms of capitalist critique in this context: while social critique attacks the exploitative character of capitalism, artistic critique focuses on the domination of capital and alienated labour (Boltanski and Chiapello 2001; Nies and Sauer 2012). Capitalism has now absorbed the critique of capitalism, which climaxed in 1968 in strikes and protests and threatened to deprive capitalism of its legitimacy, and has implemented artistic demands for autonomy, flexibility, creativity, and self-design in the form of new corporate strategies (Boltanski and Chiapello 2001; Deutschmann 2008). However, it was precisely these developments that removed the fundament for social crit- icism, since the new, decentralized, individualized and flexibilized company organization offered hardly any basis for solidarity and collective bargaining and, moreover, was once again receiving increased support from employees (Boltanski and Chiapello 2001). This chapter adds three theoretical approaches to the study. Firstly, Boltanski and Chiapello interpret the increasing spread of unbounded, flexibilized, and subjectified forms of work organization as a pure reaction of capitalism to the crises caused by the increasing criticism of capitalism and as a strategy to overcome these crises and ‘to involve workers again in the productive process’ (Boltanski and Chiapello 2001, 469). In fact, rather the convergence of a globalizing capitalism and the increasing demands of workers for autonomy, 3 Max Weber uses the term of the ‘spirit of capitalism’ to describe a certain way of life and attitude to work which, from a context of origin foreign to capitalism—namely the Calvinist doctrine of predestination—drove the emergence of capitalism through the social practice of the actors equipped with this ‘spirit’ (see Weber, 1965[1920]). 152 flexibility, and subjectivity can be assumed (Deutschmann 2008; Nies and Sauer 2012). Under the conditions of expanding global financial market capi- talism, the previously established bureaucratic and hierarchical organization of work within self-contained companies, but also the associated social security systems, which were intended to protect workers from direct access to the market, ultimately proved to be an inhibiting factor for productivity. These included, at the societal level, the dismantling of social security systems and the associated re-commodification of labour (Bourdieu 1998; Castel 2009; Nies and Sauer 2012) and, at the company level, the transition from bureau- cratic-hierarchical control by the autonomous self-control of employees. The central mechanism is the implementation of self-employment in dependent employment. The dependent employees are now largely autonomous in the performance of their work, but are also—just like self-employed entrepre- neurs—directly confronted with the demands and mechanisms of the free market and the effects of their own actions. Their performance is no longer measured in terms of labour input, but in terms of output. In addition, the authors pursue the perspective that the scope for autonomy implemented in the new type of work organization are to be equated with the substantive demands of artistic criticism of capitalism (Deutschmann 2008; Nies and Sauer 2012). In fact, since the 1990s, a comprehensive transformation of the company work organization can be observed, which certainly includes new kinds of flexibility and autonomy for the employees. However, this is by no means a seamless redemption of these demands, but rather an alienated takeover under the opposite sign (Deutschmann 2008; Nies and Sauer 2012). Employees are now less the target of operational forms of management con- trol, but remain equally involved in external constraints—such as market or customer requirements—which are largely beyond their own influence. The greater flexibility and autonomy also force employees to apply a high degree of self-organization, rationalization, and discipline and to independently manage and control the transformation of their workforce into concrete work. The increased scope for autonomy is thus not to be understood as an ‘inversion of the object principle, but rather [as] the intensified and changed objectification of the subject’ (Kratzer 2002, 3), which now does not take place from the outside, but rather through the subject itself, and is finally also reflected in the internalization and application of market-economy require- ments such as rationalization, flexibility, self-discipline and optimization, as well as permanent availability. The developments outlined are thus by no means to be characterized as implemented freedom with certain negative side effects, but on the contrary must be described precisely as a ‘consequence of continuing lack of freedom’ 153 (Nies and Sauer 2012, 42) and as a ‘change in the form of domination’ (Nies and Sauer 2012, 42) of capital over labour. Workers are given more opportunities to organize their own working hours, place of work, and work action, but they are not free to decide on them, as they remain bound by constraints and now take on additional control and monitoring tasks that were previously carried out by the company or management (Kratzer 2002; Moldaschl and Sauer 2000). However, it is precisely these connections between autonomy and dom- ination that usually remain concealed for the employees. With Boltanski and Chiapello (2001), it can be assumed that capitalism must also contain a cultural dimension—an ideology—in addition to its economic dimensions. Flexibility, autonomy, and a high degree of self-organization can be part of such an ideology and contribute to the subjects experiencing themselves as self-effective and independent in their actions, while they anticipate the ab- sence of directly experienced operational control as freedom from domination in general. This, in turn, could lead them not to reflect on the more abstract domination by processes and constraints that have become independent— and not to reflect on supervisors and operational guidelines that cannot be concretely experienced—but to perceive their claims to flexibility, autonomy, and self-determination as already realized in their work (Eichler 2005; Nies and Sauer 2012). The extent to which the increased scope for flexibility and autonomy in the sphere of employment actually leads the subjects to experience their activity as self-determined, to promote their ‘own exploitation’, and to deny existing power relations is ultimately a question that needs to be empirically supported. The underlying study intends to contribute to this by conceiving work on digital platforms as a field in which the developments described culminate in a particularly drastic way. Work on digital platforms may offer a high degree of flexibility and autonomy to the people working there, since they can determine their place of work, their working hours and—in contrast to dependent employees—also the amount of work themselves. At the same time, they hardly have security and must bear all occupational risks themselves (see section 2.3). It can therefore be assumed that the flexibility that working on digital platforms offers employees has an ambivalent character, which is made tangible by the theoretical framework outlined above: On the one hand, there are the actually existing new degrees of freedom, which can be used productively by the people working there; on the other hand, there are independent constraints which crowd workers face. It could also be assumed that the high degree of flexibility on digital platforms also requires a high 154 degree of self-organization and self-discipline, since the income is directly linked to one’s own performance. To what extent the subjects deal with this ambivalence will be shown in the empirical part of this chapter. 7.4 Research Design The underlying study4 aims to reconstruct the frames of orientations which structure the actions of crowd workers on digital platforms in order to find out what (new) demands they encounter there, whether and how they use them productively for themselves, and how they cope with the negative as- pects of platform work. The ambivalent character of the scope for flexibility and autonomy which the activities on crowd work platforms offer is the leading focus of the research. The main question of the study is, therefore: How do crowd workers deal with the ambivalent character of flexibility on crowd work platforms? In order to answer this question, it must first be clar- ified what flexibility actually means for the subjects investigated and what demands they have on their platform activities. On this basis, it can then be analysed whether and how the subjects investigated can productively use the possibilities of platform work for themselves and/or ultimately adapt themselves to the activity. In order to empirically investigate this epistemological interest, a qualita- tive, single-case-oriented, comparative research design is used. For this reason, semi-structured interviews with narrative components with four crowd workers on two German platforms were initially conducted. The Documentary Method was chosen for the analysis of the collected data. This method is specifically focused on the reconstruction of action-structuring knowledge, which, as described at the beginning, is the focus of the underlying study (Przyborski and Wohlrab-Sahr 2014). Although (theoretically) generalisable results can- not be obtained from the very small number of cases, the case-comparative analysis allows the identification of structures relevant to all cases and the elaboration of the latter into more abstract theoretical concepts, which is the aim of the underlying study. Digital platforms have been selected as a relevant research field, where clients are provided with work orders by means of an open call to contractors who are registered on the respective platform and are then self-selectively accepted by the latter and processed in return for monetary remuneration. A 4 The underlying study took place between 2017 and 2018 and was part of a master thesis submitted in 2018. For the entire study, see Gajewski (2018). 155 further selection criterion was the complete processing of the work orders via the digital platform, since the possibility of working entirely in virtual space and the anonymous and standardized communication between the client and the contractor open up completely new temporal and spatial flexibility and autonomy, which constitute the special character of digital platform work (Przyborski and Wohlrab-Sahr 2014). The sample used for the underlying study comprises a total of four cases of two crowd work platforms based in Germany. The platforms were contacted by email and asked for support in the search for interview partners. They published the interview request so that the crowd workers registered there could take part in the study. At the beginning of the study, interest in the findings was still quite openly directed at the (experienced) working conditions on digital platforms. In the course of the evaluation, the concept of flexibility emerged as particularly relevant in all cases, so that the interest in the findings was intensified in this respect. A detailed examination of the respondents’ orientations with regard to the concept of flexibility revealed the ambivalent character of the latter in the experience of the respondents. In accordance with the usual procedure in qualitative research designs, the question was developed from the empirical material in a circular process of initially relatively open survey, initial evaluation, and readjustment of the interest in knowledge (Przyborski and Wohlrab-Sahr 2014). 7.5 Empirical Results In this section, the empirical results5 of the qualitative study are presented and analysed with regard to the formulated interest in knowledge and in- cluding the theoretical concepts described in section 7.3. For this purpose, the individual cases were abstracted and synthesized with regard to various sub-questions. First of all, the question is examined what demands the in- terviewees have on their work on digital platforms. The second section then reconstructs the working conditions actually experienced on the respective platforms. Finally, it will be analysed how the interviewees deal with the possible contradictions between their own demands for flexibility and the experienced working conditions. 5 For the presentation of the platforms and for detailed case descriptions, see Gajewski (2018). For short case descriptions, see the appendix of this chapter. 156 7.5.1 Notions of Flexibility All of the interviewees come up with the concept of flexibility without being asked, which already shows that flexibility is apparently very relevant for the crowd workers interviewed. Flexibility for them is a positively connotated term for certain working conditions. Also, flexibility is not primarily interpreted as a requirement, but as a concession of the activity. An exception is Simon, who reflects on the dual character of flexibility and recognizes that it can also be instrumentalized in order to provide more comprehensive access to the workforce. In all four cases, however, the findings are consistent in that flexibility reflects the central demand of the respondents for their work and is also fully realized in the platform activity. But what does the term flexibility mean for the respondents? The analysis of the available interview data shows that, with the exception of Nicolai, all of the interviewees use the platform activity to earn an addi- tional income in addition to their main (gainful) occupation.6 In the cases Simon, Julia, and Lukas, a time-consuming and money-consuming private interest (Simon and Julia) or a special life situation (Lukas) is added. This gives rise to the claim of the interviewees that the platform activity should be compatible as smoothly as possible with the other life context. This, in turn, requires the highest possible degree of independence in terms of time and space, the absence of external control, and the ability to determine the place of work and working hours themselves. The interviewees actually find these factors in their respective platform activities. For this reason, most of the respondents perceive their own flexibility requirements as already fulfilled (Gajewski 2018). As the explanations in section 7.3 made clear, spatial and temporal flexibility and low operational management and control mechanisms in combination with a higher degree of self-organization are not a genuine characteristic of platform work, but have been spreading throughout the entire labour market for decades. For this reason, it is a noteworthy finding that activity on digital platforms, due to the virtually non-existent contact to the clients and the fact that the work is carried out in a completely self-organized manner on one’s own, private digital work equipment, sometimes opens up completely new types of flexibility, which are used by the interviewees in order to open up life 6 Simon works on the platform alongside full-time studies and Lukas works mainly as a tour guide and generates an additional income through the platform. Julia earns–– apart from other sources of income within her household––their main income on the platform, but they do voluntary work, which is comparable to gainful employment. Nicolai is currently looking for work and uses the platform income as additional income to his unemployment benefit. 157 opportunities which could otherwise not be realized. Thus, this form of activ- ity makes it possible to work from almost anywhere in the world—assuming there is internet access—and in combination with the sometimes quite high attainable income also the necessary financial resources for a global lifestyle:7 I told you I was a little annoyed by this article [in a magazine] […] it said that the only comfort of these people is that they can do their work on the beach of uh Thailand. No. Then I said ok, right, I’ve done that often in Thailand, that’s true, but that’s belittling. So, I say, I have seen things already, Cambodia, Angkor Wat, I […] have seen things I probably would not have seen my whole life. Well, as a student I would not have been able to finance it, nor would I have been able to find the time […] I have to say that clearly (Simon). A further potential that the platform activity offers, and which is particularly important in the case of Simon, is the possibility of being able to control one’s own profits on the platform by means of one’s own performance, instead of receiving a fixed income independent of one’s own productivity. The faster the individual orders are processed, the more orders can be accepted, which can significantly increase the income. The fact that work is not done for a single employer, but for many anonymous clients, allows subversive strategies to disguise the actual speed of work in order to control prices on the platform market. This can be seen in the following quotation: You can use this optimization at [Platform 2] for yourself. I don’t think they know how fast I write the texts. I think they have no idea […] And I’m also trying hard that they don’t get it right. So, I suppose if I accept a text now, I don’t write it, and um then enter it, because then they would see that there are about 5 minutes be- tween accepting and handing in the text, the system records that, but I do some other one and then do it so that it can no longer be traced (Simon). On the whole, it can be summarized that working on digital platforms opens up new kinds of flexibility and autonomy for the subjects investigated, which would not be found either in flexibly organized employment relationships or in ‘traditional’ self-employment, and which the interviewees also experience positively. Whether and to what extent the respondents’ demands for flexi- bility also contradict the working conditions on the digital platforms will be examined in the next section. 7 The following interview quotes have been translated as accurately as possible from the German original into English. 158 7.5.2 Flexibility Requirements on Crowd Work Platforms A striking contradiction, which can be identified in all four cases, is that the subjects investigated insist on their flexibility and autonomy in the perfor- mance of their activities on the platforms, but at the same time numerous factors can be reconstructed on the basis of the narratives and descriptions from their everyday working lives, which are diametrically opposed to these claims, but which are not reflected by the interviewees as restrictions of their own claims to flexibility and autonomy. After all, it is not a company framework such as fixed working hours, the necessary presence at the work- place, or the obligation to follow the instructions of a superior to which the respondents have to adapt. Nevertheless, it is apparent that they are not able to adapt their platform activities flexibly to their own life situation and current needs. Assuming that they are dependent on achieving a relatively stable monthly income from the platform activity, which is the case in all the cases selected here, they must adapt their time management flexibly to the platform activity and thus ultimately to market economy factors. They are initially dependent on the order situation, which may well fluctuate. This means that the examined crowd workers can—and must—work on the platform whenever there are orders. This situation is further aggravated by the strong competition with other crowd workers who also depend on the orders. The most orders—and therefore, the biggest profit—is made by those who accept the orders the fastest and then process them the fastest in order to be able to accept further orders. Once an order has been accepted, it must be completed within a specified period of time, regardless of whether it requires work at night, on weekends or on public holidays. If deadlines are not met, this has an immediate and long-term impact on the own reputation, which in turn means more difficult access to orders. So while the platform activity offers the respondents a certain amount of flexibility, it also demands it from them in the first instance (Gajewski 2018). This applies in particular to the cases Lukas and Nicolai, who do not have a special relevant qualification, but also have few opportunities for advance- ment within the platform, since they process standardized microtask orders and are therefore in a special way dependent on the respective order situation. Simon and especially Julia, on the other hand, have partially succeeded in emancipating themselves from current fluctuations in the number of jobs on offer and competition with other crowd workers for jobs on the platform, but only on the basis of a status on the platform (Julia) that is privileged by their own professional training and consistent, above-average performance. 159 Due to the easy, fast, and flexible access to an almost unlimited volume of tasks that the platforms open up, the value of the individual’s work is permanently threatened by cheaper and better available labour, which is especially evident in the case of Julia. Activities which in the past required specific qualifications can now be performed by unskilled workers at lower prices but of similar quality via the digital platform (Gajewski 2018). 7.5.3 Coping Strategies of the Working Subjects The central problem that the working subjects on digital platforms have to deal with is the fact that the income is directly dependent on the amount of orders processed, which is limited on the one hand by the needs of the orderers and the competition of the other crowd workers, and on the other hand, can be controlled on the basis of one’s own work performance. The subjects of the study work on this problem by resorting to various (self-)rationalization strategies. The procedure of order processing itself is first of all—especially in cases Simon and Lukas, partly also in case Julia—permanently aimed at increasing the own working speed in order to be able to process more and more orders in ever shorter time. This is accompanied by self-developed strat- egies to minimize the amount of work involved in processing orders, which sometimes deliberately undermine the regulations of the platform and/or the specifications of the client. In the direct confrontation with competition and the dependency of profit on productivity, the persons examined thus make themselves the object of measures of work intensification: So, when you do the job for the first time, it’s still like, yes, you still think. At some point you stop thinking. Uh preferred page, click on it once and it does that automatically and uh of course the moment the page loads, you already upload the next job for example. So […] you develop your tricks to make the whole thing faster (Lukas). A contrasting practice can be found in the case of Julia, who due to her high qualifications no longer competes directly with other contractors for the orders, but is indirectly threatened by the existence of many other, possibly cheaper offers. The orientation towards the permanent improvement of one’s own work performance is also clearly present in this case. The difference to the other cases is that the focus here is not on increasing output, but on improving quality. Ultimately, it can be reconstructed for all cases that they counter the pressure of competition and performance on digital platforms by internalising and applying the capitalist principle of permanent performance improvement. 160 The dependence of income on one’s own work performance, the order situation and competition with other contractors also means that the inter- viewees orientate their working hours to the platform conditions or place their lives entirely at the service of the platform activity. This is where the differences between Lukas and Nicolai, who work on platform 1, and Simon and Julia, who work on platform 2, become apparent. Lukas and Nicolai compete with many other contractors for limited microtask jobs that can be completed in seconds to minutes and therefore need to be accepted as fast as possible. For this reason, they both have built up a reliable experience of the times at which on average many jobs are put online most often and align their working hours with these times. In comparison, Simon and Julia do not compete so strongly for orders, but here too the mechanisms of unrestricted work are effective. Both of them have a strong tendency to measure the nec- essary reproduction phases by the amount of work lost on the platform or even to leave them out in favour of work, either to be able to process more orders and thus maximize one’s own profit (Simon) or to process the existing orders much more carefully than necessary in order to consolidate one’s own status (Julia) (Gajewski 2018). In all cases, it can be seen that the subjects investigated do not adapt their working hours to their living conditions in a self-determined way, but on the contrary subordinate their living conditions to their platform activities. Interestingly, it is shown that all the subjects studied, despite the emphasis on the possibility of flexible working hours, follow a fixed daily structure with regular working hours. This can particularly be shown in the cases Simon and Julia, who succeeded in emancipating themselves from the current order situation and the competition with other crowd workers on the platform, whereas Lukas and Nicolai have to orient themselves more towards these two factors, even though a fixed structure is in principle an important reference point for their working time organization. In the case of Simon, it can be shown that this frame of orientation, which can be observed in all cases, represents a protective function against the extended access of the platform to the interviewees’ life and serves to draw the line between the spheres of work and reproduction, which are in principle delimited by the possibilities of digital platforms. 7.5.4 Synthesis: Ambivalent Flexibility on Digital Platforms What is the ambivalent character of flexibility on digital work platforms, and how do the subjects working there deal with the new kinds of flexibility, 161 but also the pressure of competition and performance? This question will be answered in the following on the basis of the findings to date. Work on digital platforms is carried out in a self-organized manner and enables—especially on platforms where more complex, long-term, and qual- ified activities are mediated, as is the case for platform 2—a comparatively high income, which can be controlled by the own productivity and subver- sive strategies unnoticed by the clients. These circumstances lead to the fact that working on digital platforms can, at least under certain circumstances, actually open up new kinds of flexibility, which have not been offered in this form by any other form of employment so far. The fact that the positive aspects of flexibility come into play in the cases Simon and Julia is, however, due to a number of factors that cannot be generalized for all crowd workers and not for all types of platforms. Both of them have been working on the platform for a long time, and they were able to acquire a high status on the platform through continuous and committed performance, from which they can now profit in the form of many and well-paid jobs. The fact that the scarcity of orders in particular means that work on the digital platform is not flexible at all is shown, on the other hand, by the cases of Lukas and especially Nicolai, who have to make their time management dependent on the current order situation and even then only make small profits that are only useful as additional income. This already shows the ambivalent character of flexibility on digital platforms. For the subjects can, at least in some cases, only determine their working hours, their scope of work, and their target income relatively flexibly because they are directly linked to the mechanisms of the free market without any company and welfare state protective function and their performance is measured exclusively on the basis of the result of work and no longer on the expenditure of labour. The same factors which at least in some cases lead to the fact that the subjects on digital platforms can flexibly determine their working time, their scope of work, and their target income, thus restrict any form of self-determination in the same time, since they entangle the subjects in independent constraints—in this case, the limited availability of orders and the competition with other crowd workers—which they themselves cannot influence, while at the same time they are directly confronted with the consequences of both these external conditions and their own actions. In addition to increased flexibility, crowd workers are thus primarily confronted with imperatives of flexibility, autonomy, and self-organization (Nies and Sauer 2012). The findings of the underlying study also show that the persons inves- tigated react to the extended and flexibilized access to themselves and their 162 life context with self-rationalization. On the one hand, the platforms allow working hours around the clock and a large potential work volume, but on the other hand, they make both of these absolutely necessary due to the low income. In order to at least partially prevent this access to their entire lives, the interviewees—especially Simon, Julia and Lukas—make themselves objects of self-applied labour intensity by constantly increasing their work rate and the output produced. In addition, their entire lifetime becomes po- tential working time, so that reproductive phases are only assessed in terms of missed productivity. An important finding of this study shows that the interviewees nevertheless insist firmly on the positive aspects of their platform activities. The flexibility becomes an instrument for domination and control in two ways: on the one hand, it demands a flexible adaptation of the subjects’ entire life context to the conditions on the platform as well as a high willingness to perform and a permanent increase in productivity; on the other hand, it motivates them to continuously participate in the capitalist production process through the illusion of already redeemed self-determination. Ultimately it could be shown that work on digital platforms produces an ambivalent form of flexibility, which opens up potential for action for the crowd workers, but ultimately remains an instrument of domination, which extends and facilitates capital’s access to work, in that this access is no longer exercised in the mode of an external compulsion, but is internalized by the subjects and applied to themselves (Bechtle and Sauer 2001; Eichler 2009; Kocyba 2005; Kratzer 2002; Nies and Sauer 2012). For this reason, however, they do not experience the exercise of domination and control as an external compulsion, but rather as self-determined action. In this way, domination is both concealed and extended (Eichler 2005; Nies and Sauer 2012). 7.6 Conclusion The aim of the underlying study was to work out the ambivalent character of the flexibility of digital platform work for the crowd workers and to reconstruct how they deal with the novel scope for flexibility, but also the requirements of this still relatively new form of work. As a first approach to the field of digital platform work showed, crowd work is a very heterogeneous object, ranging from the smallest, low-paid micro-tasks to longer-term, creative, and demanding activities. Although initial statistical findings are already available, there is a need for further research, especially with regard to the very different working conditions on the various platform types. 163 The chosen theoretical perspective places crowd work in the field of post-Tayloristic forms of work, in which the company level as a mediator between market and individual is eroding, which on the one hand means more flexibility and autonomy for the employees, but on the other hand also new requirements with regard to self-organization and availability. This perspective opened up a view both of the subjects of digital platform work and of the ambivalent effects of new forms of organization of work on the latter and considered the subject-oriented nature of the underlying study. Thus, crowd work could also be examined from a technological, organiza- tional, macro- or transnationalization sociological perspective. In particular, the landscape of theory-based empirical social research on crowd work is still almost uncharted, so that further research would be desirable. The analysis of the empirical data in this study was aimed at reconstructing the frame of reference of the interviewed crowd workers in order to work out the experienced ambivalence of the flexibility that working on digital platforms entails. It turned out that, despite the differences between the two platforms on which the respondents work, there are clear similarities in the experience of the possibilities and requirements as well as in how the latter are dealt with by all respondents. The central finding of this study is that the respondents experience their activity as flexible and self-determined, but are actually confronted with external constraints on the platform which are be- yond their own influence, but to which they must flexibly adapt themselves. Subsequently, the interviewees react with self-rationalization by permanently shortening their processing time—through routines and self-developed processing strategies—or by increasing the output per time unit in order to emancipate themselves at least to some extent from the platform boundaries. Ultimately, however, they promote their own exploitation, as they make their working capacity and lifetime available on an ever-greater scale. Under these conditions, the scope for flexibility that platform work actually offers, becomes an ideology that provides the necessary motivation for further participation in the capitalist production process and self-exploitation. Due to the small number of cases, some limitations have to be noted with regard to the scope of the results. For example, a theoretical generalization of the results beyond the cases analysed in the underlying study is not possible, because the sample is too small and was not carried out systematically. Nev- ertheless, it was possible to reconstruct structures that are effective beyond the individual case. The empirical results of this study are therefore not (yet) to be understood as an explanatory approach or a theory close to the subject matter, but rather as first concepts to be enriched and verified by means of further, systematically selected cases. 164 Further need for research recently arises from the differences between the respondents regarding the two different platforms. It could be shown that the positive aspects of the platform work are especially effective for those respon- dents who already have a qualification relevant for the work on the platform and/or who generally have higher qualifications, whereas those respondents who accept microtask assignments are more likely to be affected by the neg- ative effects. Here it would be interesting to ask whether and to what extent crowd work creates and/or reproduces new, old or ‘new old’ labour market inequalities between qualified freelancers and ‘digital workers’. 7.7 References Al-Ani, Ayad and Stefan Stumpp. 2015. Motivationen und Durchsetzung von Interessen auf kommerziellen Plattformen. Ergebnisse einer Umfrage unter Kre- ativ- und IT-Crowd workern. HIIG Discussion Paper Series No. 2015‒05. 7 De- cember, 2015, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2699065 (accessed March 11, 2021). Bechtle, Günter and Dieter Sauer. 2001. Fordismus als Zwischenspiel. Zur he- terogenen und ambivalenten Entwicklung des gegenwärtigen Kapitalismus. In Hilde Wagner (Ed.), Interventionen wider den Zeitgeist. Hamburg: VSA. Benner, Christiane. 2016. Crowdsourcing: Die radikalste Form der Digitalisierung von Arbeit. Eine Herausforderung für gewerkschaftliches Handeln. In Lothar Schröder and Hans-Jürgen Urban (Eds.), Gute Arbeit. Digitale Arbeitswelt – Trends und Anforderungen. Frankfurt am Main: Bund-Verlag. Bertschek, Irene, Jörg Ohnemus, and Steffen Viete. 2015. Befragung zum sozio- ökonomischen Hintergrund und zu den Motiven von Crowd workern. Endbericht zur Kurzexpertise. Mannheim: Zentrum für Europäische Wirtschaftsforschung (ZEW). Boltanski, Luc and Ève Chiapello. 2003. Der neue Geist des Kapitalismus. Kon- stanz: UVK. Boltanski, Luc and Ève Chiapello. 2001. Die Rolle der Kritik in der Dynamik des Kapitalismus und der normative Wandel. Berliner Journal für Soziologie, 11(4): 459‒477. Bourdieu, Pierre. 1998. Gegenfeuer. Wortmeldungen im Dienste des Widerstands gegen die neoliberale Invasion. Konstanz: UVK. Castel, Robert. 2009. Die Wiederkehr der sozialen Unsicherheit. In Robert Castel and Klaus Dörre (Eds.), Prekarität, Abstieg, Ausgrenzung. Die soziale Frage am Beginn des 21. Jahrhunderts. Frankfurt and New York: Campus. Däubler, Wolfgang and Thomas Klebe. 2015. Crowdwork: Die neue Form der Arbeit. Arbeitgeber auf der Flucht? NZA, 2015: 1032‒1041. 165 Deutschmann, Christoph. 2008. ‘Kapitalismus’ und ‘Geist des Kapitalismus’ – Anmerkungen zum theoretischen Ansatz von Boltanski/Chiapellos. In Gab- riele Wagner and Philipp Hessinger (Eds.), Ein neuer Geist des Kapitalismus? Paradoxien und Ambivalenzen der Netzwerkökonomie. Wiesbaden: VS. Durward, David, Ivo Blohm, and Jan Marco Leimeister. 2016. Crowd Work. Business and Information Systems Engineering, 4(58): 1‒6. Eichler, Lutz. 2009. Dialektik der flexiblen Subjektivität. Beitrag zur Sozialcha- rakterologie des Postfordismus. In Stefan Müller (Ed.), Probleme der Dialektik heute. Wiesbaden: VS. Eichler, Lutz. 2005. Seelenlose Unternehmer ihrer Selbst? Über die psychische Wirkung subjektivierter Arbeit. In Arbeitsgruppe SubArO (Eds.), Ökonomie der Subjektivität – Subjektivität der Ökonomie. Berlin: Edition Sigma. Eurofound. 2015. New Forms of Employment. Luxembourg: Publications Office of the European Union. Felstiner, Alek. 2011. Working the Crowd. Employment and Labor Law in the Crowdsourcing Industry. Berkeley Journal of Employment and Labor Law, 32(1): 143‒203. Gajewski, Eltje. 2018. Ambivalente Flexibilität: Crowd workerinnen zwischen selbstbestimmter Arbeit und kapitalistischer (Selbst-)Verwertung auf digita- len Plattformen. Duisburger Beiträge zur soziologischen Forschung, 2018‒03. https://www.uni-due.de/soziologie/duisburger_beitraege_dbsf-2018-03.php, (accessed March 11, 2021). Gerber, Christine and Martin Krzywdzinski. 2019. ‘Brave New Digital Work? New Forms of Performance Control in Crowdwork.’ In Steve P. Vallas and Anne Kovalainen (Eds.), Work and Labor in the Digital Age (Research in the Sociology of Work), 33: 121‒143. Kenney, Martin and John Zysman. 2019. ‘Work and Value Creation in the Platform Economy.’ In Steve P. Vallas and Anne Kovalainen (Eds.), Work and Labor in the Digital Age (Research in the Sociology of Work), 33: 13‒41. Kocyba, Hermann. 2005. Selbstverwirklichungszwänge und neue Unterwerfungs- formen. Paradoxien der Kapitalismuskritik. In Arbeitsgruppe SubArO (Eds.), Ökonomie der Subjektivität – Subjektivität der Ökonomie. Berlin: Edition Sigma. Kratzer, Nick. 2002. Warum machen die Beschäftigten mit? E rklärungsansätze für die Ambivalenz selbst organisierter Arbeit. München: Institut für Sozial- wissenschaftliche Forschung e.V. ISF München. Kuek, Siou Chew, Cecilia Maria Paradi-Guilford, Toks Fayomi, Saori Imaizumi, and Panos Ipeirotis. 2015. The Global Opportunity in Online Outsourcing. Washington, D.C.: World Bank Group. Leimeister, Jan Marco, David Durward, and Shkodran Zogaj. 2016. Crowd wor- ker in Deutschland. Eine empirische Studie zum Arbeitsumfeld auf externen Crowdsourcing-Plattformen No. 323. Hans-Böckler-Stiftung: Düsseldorf, https://www.boeckler.de/pdf/p_study_hbs_323.pdf (accessed March 11, 2021) 166 Leimeister, Jan Marco and Shkodran Zogaj. 2013. Neue Arbeitsorganisation durch Crowdsourcing. Eine Literaturstudie. Düsseldorf: Hans-Böckler-Stiftung. Menz, Wolfgang and Ana Cárdenas Tomazic. 2017. Gerechte neue Arbeitswelt? Crowd working aus normativ-empirischer Perspektive der Soziologie. In Martin Dabrowski and Judith Wolf (Eds.), Crowd working und Gerechtigkeit auf dem Arbeitsmarkt. Paderborn: Schöningh. Moldaschl, Manfred and Dieter Sauer. 2000. Internalisierung des Marktes – Zur neuen Dialektik von Kooperation und Herrschaft. In Heiner Minssen (Ed.), Begrenzte Entgrenzungen. Wandlungen von Organisation und Arbeit. Berlin: Edition Sigma. Nies, Sarah and Dieter Sauer. 2015. Arbeit – mehr als Beschäftigung? Zur arbeits- soziologischen Kapitalismuskritik. In Klaus Dörre, Dieter Sauer and Volker Wittke (Eds.), Kapitalismustheorie und Arbeit. Neue Ansätze soziologischer Kritik. Frankfurt and New York: Campus. Pongratz, Hans Jürgen and Sarah Bohrmann. 2017. Online-Arbeit auf Internet- Plattformen. Empirische Befunde zum ‚Crowd working’ in Deutschland. Arbeits- und Industriesoziologische Studien, 10(2): 158‒181. Przyborski, Aglaja and Monika Wohlrab-Sahr. 2014. Forschungsdesigns für die qualitative Sozialforschung. In Nina Baur and Jörg Blasius (Eds.), Handbuch Methoden der empirischen Sozialforschung. Wiesbaden: VS. Schörpf, Philip, Jörg Flecker, Annika Schönauer, and Hubert Eichmann. 2017. Triangular love-hate: management and control in creative crowd working. New Technology, Work and Employment, 32(1): 43‒58. Stanford, Jim. 2017. The Resurgence of Gig Work: Historical and Theoretical Perspectives. The Economic and Labour Relations Review, 28(3): 382‒401. Weber, Max. 1965 [1920]. Die protestantische Ethik. Eine Aufsatzsammlung. Her- ausgegeben von Johannes Winckelmann. München and Hamburg: Siebenstern. Appendix: Case description Simon is male, in his early 30s, is writing his Master’s thesis in an engineering degree programme at a German university at the time of the interview, and has been working as a copywriter on Platform 2 since 2011 (i.e. for about six years at the time of the interview) to earn a living alongside his studies. According to his own information, he now earns about € 1,500 to € 1,600 a month, for which he works three to four hours a day. Lukas is male, forty-two years old, and lives with his partner and child in Brazil, where he emigrated from Germany in 2010. There, at the time of the interview, he works full-time as a tour guide and part-time as a micro jobber, mainly in the area of data maintenance, on Platform 1. He works there about once a week for several hours and generates a monthly income of 167 about € 100, whereby this is worth about four times as much as in Germany due to the weak Brazilian currency. Julia is female, about fifty to sixty years old, lives in Germany, and has been working on Platform 2 as a copywriter for several years (no exact details). The platform job is her main occupation and she does not do any other jobs; she achieves a monthly income of € 1,600 to € 2,000. An important part of her everyday life is her voluntary rescue dog work, which takes up several hours of her time every day. Nicolai is male, in his mid-40s, and lives in Bosnia-Herzegovina, where he was also born. Since 2014/2015, he has been working on Platform 1, as according to him, this is the only platform where jobs for Bosnia and Herze- govina are available. There he accepts assignments in the field of microtask, more precisely in the field of data maintenance. There is no information on the exact working hours and duration of the work, as there is only a very limited number of jobs available for Bosnia-Herzegovina and Nicolai always accepts as many jobs as possible. He uses the monthly income of about € 50 earned in this way as a supplementary income in addition to his unemployment benefit. Part IV Work at the Margins 171 8 The Role of Work Before and After Retirement on Poverty Dynamics in Old Age. Evidence from a Follow-Up Study in Switzerland Rainer Gabriel, Michel Oris, Sonja Kubat, Kushtrim Adili, and Monika Götzö 8.1 Introduction As in many other European countries, the population in Switzerland is age- ing. With baby-boom cohorts entering retirement, the share of people aged sixty-five and older has reached 18.7 percent in 2019 (Swiss Federal Statistical Office 2020). Together with declining returns from financial investments, this demographic shift has placed considerable stress on the pension system. However, with the discussions primarily focusing on the financial sustain- ability of the system, the questions of financial security and poverty in old age have been pushed into the background. This is happening even though old age poverty (OAP) remains relatively prevalent among the population aged sixty-five and older in Switzerland, with estimations for income-poverty around 16 to 20 percent (Gabriel et al. 2015; Guggisberg and Häni 2014; Oris et al. 2017; Wanner and Gabadinho 2008). In addition, OAP remains insufficiently understood. This is problematic given that substantial adjust- ments to the pension system are being discussed and could be implemented in the coming years. Given this societal and political background, and in the context of this book, our principal aim in this chapter is to synthesize what is currently known on the link between work and OAP and to then offer a discussion on how this nexus might develop in the future. For this investigation, the most important starting point lies in the convergence of existing studies on OAP towards the finding that people’s previous profession—no matter whether this is measured at the beginning of their work-life or at the end—appears to be strongly linked to the risk of 172 experiencing financial hardship in old age (Oris et al. 2017). Interestingly, work-trajectories (meaning the sequence of years spent either working or in unemployment) do not explain this relationship: Hence, for instance, manual labourers are not more susceptible to OAP because they might have experienced more episodes of unemployment and job changes (Gabriel et al. 2015). This finding has been interpreted as evidence for a social stratification framework, meaning that it is, above all, a person’s position in the economic structure—whereas people’s job can be considered a proxy for the latter—that determines a wide range of living conditions, and in our case the financial situation in old age. Moreover, the occurrence of critical life events, which for the whole of the EU have been proven to have a poverty-triggering effect (Vandecasteele 2011) have not been found to play a significant role in the Swiss context (Gabriel et al. 2015). Beyond these well-established findings, numerous questions about OAP remain unanswered or inconclusive: Above all, the clear causal mechanisms and pathways for this pattern remain inconclusive. Neither is there clear evidence regarding the validity of social stratification for other measures of poverty that are not income-based. Also, little is known on what effect working beyond the legal retirement age has on poverty. And finally, despite the fact that poverty has been shown to contain an important temporal dimension (in the population of working age, only very few seem to remain in a situa- tion of poverty for a prolonged time, but there are considerable movements into and out of poverty), OAP has almost never been assessed using such a dynamic view. Therefore, keeping this chapter’s overarching aim in mind, the empirical part of our chapter addresses three aspects which we believe currently remain inconclusive in the existing literature: The first relates to the role of work after retirement on OAP, the second focuses on how poverty evolves over time, and thirdly, we explore the question whether the findings from income-poverty can be applied to the subjective perception of people’s financial situation. Considering the dominant role that social stratification has played for previous analyses of OAP, we once again refer to this theoretical framework for our analyses and derive all of our hypotheses from it. In the following part, we discuss the existing literature on OAP in more detail, providing an overview of the international literature and research in Switzerland and placing particular emphasis on the aspects of temporal dy- namics, poverty-measures, as well as the role of work. We then build on our main theoretical background of social stratification and critical gerontology to derive our working hypotheses. The third part presents data and methods. The fourth section summarizes the results and part five discusses them. In this fifth and final part, we will also provide our view on how this relationship 173 might evolve in the future—based on the supposed transformations of work that are described in this book. In that sense, we attempt to ‘look back to look ahead’ (Kohli 2007). 8.2 State of the Art: Research on Poverty in Old Age 8.2.1 Three Historical Phases Of Poverty Research Oris et al. (2017) suggest that poverty1 research can be divided into three distinct phases. The first being the pre-welfare state phase, the old regime of 19th century industrial society where poverty was basically the norm for the majority of the working class (Rowntree 1901). Then followed the phase of welfare-state expansion around the middle of the 20th century, where poverty reduction and prevention were key concerns for public policies. The third phase could be entitled as post-welfare state transition, with globalization, (post-)modernization and population ageing increasingly putting the ‘tradi- tional’ welfare state under pressure. During the first phase, the focus on poverty in old age was largely absent. This absence can above all be explained by the dominant demographic reality. The share of elderly people was much lower at the beginning than towards the end of the 20th century, and the ageing of the population was no issue at that time (Gavrilov and Heuveline 2003; Hayward and Zhang 2001). As the century progressed, the share of elderly steadily increased and gave rise to old age as an object of study in the 1970s and 1980s (Laslett 1987; Neugarten 1970). Simply put: ‘old age is relatively young’ (Samuel 2017, 153), and as a consequence, research on poverty in old age is even younger. However, this does not mean that the elderly population was not matter of concern during the past century—on the contrary: The pauperization of the elderly was a significant social problem at the end of the 19th and in the first half of the 20th century, both in Switzerland as well as in the rest of Europe, and the general public concern regarding this issue was in fact central to push welfare states to develop pension systems. In the second phase, with welfare states expanding, research on poverty intensified in order to provide empirical evidence for the design of new social policies. During this phase, the paradigm of studying poverty also started to 1 Poverty can be conceptualized in multiple ways. For a discussion of the different measures and definitions, see Haveman (2001). Almost all studies reviewed in this section utilize an absolute income poverty approach. In our own analyses we employ both an absolute income-poverty approach as well as a subjective poverty measure. 174 shift towards a more dynamic understanding (Bane and Ellwood 1983). Yet, due to the lack of longitudinal data this new approach would not result in a notable rise in corresponding studies. Insights on poverty in old age often appeared, if at all, merely as a by-product of more general poverty studies. OAP, therefore, was often presented as being static and structural—the idea being that once people rely on a form of pension, their economic situation would not change much anymore (Walker et al. 1984). It has to be noted that some authors were ahead of their time and focused on poverty in old age, such as Holden et al. (1986) who relied on a small sample of elderly couples and widows, to conclude that poverty in old age might perhaps not be static after all. Finally, the welfare states of the post-war years started to come under pressure as early as the 1970s, fueled above all by globalization, slowing eco- nomic growth but also in the context of (neo)liberal public policies (Lessenich 2015; Harvey 2011). This development intensified towards the 1980s and was at its most pronounced towards the 1990s. The rise of this post-welfare state phase also coincided with a rise in awareness of the new demographic reality of population ageing in the Western World. This sparked an ongoing debate on the future of pension systems across Europe (Bonoli 2003; Kohli and Arza 2011; Schulz and Borowski 2006) and as a consequence, towards the turn of the century numerous studies launched with the aim of better understanding the living conditions of the retired population. At the same time, influenced by the life course approach, research contributed to the rising awareness that old age is a particular life stage with regards to poverty (Leisering and Leibfried 2001). However, OAP mostly continued to be a side-topic in more general poverty studies. Also, the paradigm of poverty as a time-dependent phenomenon was increasingly incorporated into the study designs (Layte and Whelan 2003). Following the pattern that research on OAP seems to have a lag of about a decade in terms of methodology, this shift towards a dynamic view is only starting to happen today. 8.2.2 International Research on Old Age Poverty and Evidence from Switzerland As has been described so far, research on poverty in old age is a relatively young field. In addition to population ageing only stepping into the public and scientific spotlight, another reason why studies have been lacking is the absence of reliable data. Several large national projects, most of them based on large-scale surveys, around the 1990s addressed this shortcoming and have led to corresponding studies on OAP: Hauser and Neumann (1992) focused 175 on Germany, Price (2006) on the UK, Paugam (1991) on France, and Leu et al. (1997) as well as Lalive d’Epinay (2000) on Switzerland. All of these studies reached similar conclusions: poverty was found to be a significant phenomenon affecting a large share of pensioners, and gaps in the pension system were believed to be the most likely cause. Later, the Survey of Health and Retirement in Europe (SHARE) provided a first pan-European perspec- tive. And again, the conclusion pointed towards it being an underestimated phenomenon (Börsch-Supan et al. 2008). In the early 2000s, due to its de- sign as a panel study, SHARE also offered the basis for first analyses on the temporal dynamics of OAP. Accordingly, OAP was by no means found to be stable or structural, but considerable movements into and out of it were detected (Börsch-Supan et al. 2008). Surprisingly, the latter finding seems not to have been followed up by further research—not by the SHARE con- sortium, nor by other researchers. In fact, over the last decade, the emphasis on old age in the larger context of poverty studies has even grown weaker again. An important explanatory factor for this situation is that longitudinal and individual-level data remain exceedingly scarce. Instead, research seems to have gone back to more general poverty studies which does often feature old age as one area of interest, but more often place their main focus on the life course and life events as triggering mechanisms (Sandoval et al. 2009; Vandecasteele 2010; Vandecasteele 2011). As a result, research on OAP remains a field that can still be considered marginal, which led Kwan and Walsh (2018) to state in a recent literature review, the ‘focus on OAP [among existing studies] is limited’ (Kwan and Walsh 2018, 1). However, while few in number and largely focused on estimating the prevalence of poverty and exploring the associated protective and risk factors, the findings of these studies drew a mostly coherent picture. Firstly, with regards to prevalence, they found that OAP still appears as a phenomenon of substantial relevance, particularly income-poverty (Börsch-Supan 2014; Börsch-Supan et al. 2008; Haitz 2015). This also applies to Switzerland, where income-poverty at the age of retirement might even be more prevalent than in most of Europe (Guggisberg and Häni 2014; Wanner and Gabadinho 2008).2 Secondly, the studies focusing on the identification of associated risk-factors 2 The high prevalence of income-poverty is sometimes relativized, particularly in Switzerland, by pointing at the fact that retired citizens often have considerable wealth assets (see Guggisberg and Häni 2014). However, to date it has not been convincingly demonstrated—using individual level data—whether the latter means that income-poverty can be considered less problematic or might even be dismissed on these grounds. For example, owning real estate objects such as an apartment might account for considerable assets in terms of wealth but cannot easily be transformed into an actual revenue to be utilized for daily expenses. 176 also yielded coherent results relating to age/cohorts (Berthoud et al. 2009), gender (Haitz 2015; Smeeding and Sandstrom 2005; Vlachantoni 2012; Yang 2011), work status (Yang 2011), socio-economic position (Arent and Nagl 2010; Nolan and Marx 2009; Oris et al. 2017), and migration (Kaeser and Zufferey 2015; Phua et al. 2007). The impact of employment trajectories (Budowski and Suter 2002; Budowski and Tillmann 2006; Himmelreicher and Frommert 2006), as well as other life trajectories or critical life events (Gabriel et al. 2015), turned out to be less clear with contrasting results. One notable study with a more conceptual design shows OAP as the result of a complex interplay between individual factors—meaning both social and economic resources as well as having experienced certain critical life events and having experienced specific trajectories in various domains of life—and macro-political settings through pension systems (Möhring 2015). Moreover, as Kwan and Walsh (2018) also point out, most of the afore- mentioned studies are susceptible to two methodological shortcomings: an income-centred definition of poverty and the use of cross-sectional data, therefore neglecting the temporal dimension that has now become one of the cornerstones of poverty research for the general population. For Switzerland, the methodological critique by Kwan and Walsh does hold true. Accordingly, the few studies that studied OAP in Switzerland (Gabriel et al. 2015; Oris et al. 2017; Wanner and Gabadinho 2008) suffered from those limitations. Three exceptions are the qualitative study by Pilgram and Seifert (2009), the multi-focal approach by Henke (2020) cross-examining three indicators from survey data, and a report by the Swiss Federal Statistical Office comparing SILC survey data with the Swiss HABE survey, integrating multiple measures such as income, deprivation, but also accounting for wealth (Guggisberg and Häni 2014). Summarizing the literature on OAP, the few findings that appear in an almost undisputed manner is that it seems to affect a considerable part of the elderly population in European countries, Switzerland in particular, and that there seem to be distinct factors increasing or decreasing its risk. How- ever, three important questions remain and would benefit from additional evidence: First, how OAP develops over time, whether it is stable or whether there are movements in and out of OAP; second, whether the findings of the current literature which are mainly based on income-poverty, are valid for other, notably subjective, poverty-measures and third, what role work after the age of legal retirement plays. These three aspects represent the questions for our empirical contribution to the literature. 177 8.2.3 Social Stratification and Critical Gerontology Many of the previously mentioned risk or protective factors for OAP can be interpreted as being in line with a social stratification paradigm (Grusky 2001). Accordingly, it can be posited that OAP is fundamentally linked to the way society is structured with regards to socio-economic strata—or class.3 To a large extent, this link is actually the result of how pension systems—or, from a more macro-sociological perspective: whole welfare states—are designed. Switzerland, considered as a rather liberal welfare state (Esping-Andersen 1990) thus features a high level of social stratification. The theoretical framework of social stratification in OAP has been used by Oris et al. (2017) where it was argued that life course institutionalization, a term used to describe the contri- butions by people to the so-called three-pillar pension scheme in Switzerland, is a strong factor of its reproduction. The strongest dynamics appeared to be related to education, people’s first job as well as their last. More specifically, the latter relation especially applied to unskilled professions which showed a greatly increased risk of OAP. Such a view on social stratification in old age also corresponds to a critical gerontology perspective. In fact, the latter goes back to a group of researchers who applied a Marxist framework to the area of gerontology, positing that it is above all, people’s position in the economic structure during their work life, which is determinant of their living conditions in old age. According to these authors inequalities related to people’s socio-economic position are even exacerbated in old age (O’Rand 2006; Phillipson 1981; Walker 1981). Based on the outlined theoretical framework, we, therefore, derive the fol- lowing hypotheses: 1. Working after the legal retirement age is associated with increased risk of OAP (i.e. people who work at the age of retirement do so out of neces- sity) and mainly concerns people with a lower socio-economic position. 3 In the context of social stratification theory, the question of how social-economic position or class can be measured has a long tradition and is, and has always been, heavily debat- ed. In our contribution, we use work—by which we understand the type of profession a person has carried out during professional life—as a proxy for socio-economic position. In the context of this book on the future of work, our operationalization might spark the critique, that profession does not necessarily say much about the characterization of work, such as the social structure of the workplace and other aspects. While we accept this point of critique as a limitation of our contribution, we also would like to stress the fact that using a dataset for the retired population, means looking back at people who have spent the majority of their working life at a time, in the 1960s until the 1980s, where work was still much more homogenous and most of the new developments of modern or even post-modern work-conditions have, if at all, only marginally affected these cohorts. 178 2. People’s first and last job are strongly associated with temporal dynamics in OAP. 3. The subjective poverty measure is coherent with the objective measure (income poverty) and manifests the same aforementioned patterns. 8.3 Data and Methods Our choice of methodological approach relies strongly on Kwan and Walsh’s (2018) criticism regarding the shortcomings of existing research. Therefore, the present study considers multiple poverty measures by including both income-poverty and an additional measure based on individual perception of financial strain, and it considers the temporal evolution of poverty across time. 8.3.1 Dataset For our analyses we use two waves from the largest and perhaps the broadest gerontological survey among the retired population in Switzerland called ‘Vivre-Leben-Vivere’ (VLV). VLV was conducted in 2011/2012 and a follow up was carried out in 2016/2017. In order to account for the structure of the population, VLV relies on a stratified random sampling approach. The first wave included 3080 respondents, was conducted in five cantons (VS, GE, TI, BS/BL, BE), and was stratified according to sex and age-group, to allow statistically robust analyses even for the (rarer) higher age-groups (for the specific composition of the original sample see Ludwig et al. 2014 and Oris et al. 2016). The follow-up was conducted in four cantons (VS, GE, TI, BS/BL) and included 1,250 respondents, 51.5 percent of whom were male, representing roughly 40.5 percent of the original sample. The attrition between the two waves of VLV2 was mainly due to mortality (34 %), refusing to participate (18.6 %), and respondents being untraceable (17.2 %). As a result, the sample is slightly modified: For example, in VLV2, 9.4 percent of people had a low level of education compared to 16 percent in VLV1. However, the sample still contains a considerable proportion of oldest-old—in fact, the relative share of the age-group 90+ even increased somewhat to 16.6 percent. Similar patterns of attrition over six years have been observed in comparable studies among the elderly population (see Ihle et al., 2020). A critical evaluation of VLV showed that it performs well regarding minimizing survey errors and adequately representing the most vulnerable sub-population among the elderly. The latter was likely due to considerable efforts to capture the most vulnerable (Oris et al. 2016). 179 8.3.2 Target Variables The target variables are two main measures of poverty in old age: The first is a binary indicator of income-poverty based on an absolute poverty line, using answers from respondents regarding their total net monthly household income (the total financial inflow including all pension schemes but also financial returns or social benefits). It was originally a 9-level categorical scale that was adjusted to household size using the OECD modified equivalence scale meth- odology. Then, based on an absolute poverty line of CHF 2,400 per month and single person, we dichotomized the answers into a binary indicator. The choice of the threshold is based on the maximum standard first-pillar pension (the basic pension which should allow to cover basic living expenses). It is also in line with recommendations by the NGO SKOS/CSIAS, and is used by the Swiss Government to allocate social welfare or supplementary benefits (SKOS and CSIAS 2013).4 Accordingly, our weighted estimations for absolute income-poverty rates in Switzerland are 21.4 percent of 65+. As discussed in previous work (Gabriel et al. 2015), this finding based on VLV data is gen- erally coherent with other studies, albeit being generally at the higher end of the estimations. It must be mentioned that this measure for income-poverty is based on self-reported monthly household income. As such, it might be prone to response bias. Even though notoriously difficult to obtain in terms of data protection and due to a lack of a national database, a better source of data would be individual-level tax data (Wanner and Gabadinho 2008). Our second indicator is a subjective measure of poverty. It is based on the question to what extent respondents find it difficult to ‘make ends meet financially, considering their monthly incomes’. The original scale was a 4-level Likert-scale ranging from ‘very difficult’ to ‘not at all difficult’. This scale was dichotomized to create an indicator that measures whether people reported difficulties or not. Our case-weighted estimations show a prevalence of 14.7 percent who reported experiencing difficulties making ends meet. This value is slightly lower than for income-poverty. Possible explanations can lie in the fact that people might have savings that they can use for daily expenses, but other more psychological factors also act, such as frugality or 4 The SKOS/CSIAS (2013) recommendations define the threshold (a monetary amount) for the absolute poverty line in Switzerland which is required to ensure a ‘humane existence’, as guaranteed in the Swiss constitution. Their recommendations cover three areas: basic living expenses (food, clothing, transport, etc.), housing costs, and health care. Method- ologically, they are based on the real-world data on the typical expenditure of the lowest income-decile of household in Switzerland. The recommendations are updated periodically and are available online at: richtlinien.skos.ch. 180 resilience to financial hardship, both of which more pronounced in older cohorts (see Henke 2020 for an in-depth study). One limitation related to our income-poverty measure is the fact that we were not able to adjust it according to people’s wealth. The reason for this has to do with the quality of the information on wealth in the VLV questionnaire, which was assessed with a single question where respondents were supposed to indicate the cumulative amount of all wealth assets (encompassing everything from savings, shares, stock holdings, real estate ownings, etc.). Here, there seemed to be large inconsistencies on whether people included real estate holdings (i.e. owning one’s own home) in this calculation or not. Notably, a non-negligible share of individuals was found having reported low assets (< 20,000 CHF) but at the same time having also indicated owing their own home. As there was no means of reconstructing and validating the correct responses, the only solution was to completely omit this kind of information. For the longitudinal analyses we have then used the information from both waves to calculate the four possible trajectories: Whether individuals stayed stable in one or the other situation (e.g. poor in 2011 and 2016 was coded as the trajectory ‘stable poor’), or whether they changed from one into the opposite situation. 8.3.3 Covariates Three covariates are used to test our hypotheses. They can be divided into concerning people’s past work-life—which we use to proxy an indicator of social stratification—and variables on the more recent work-life within retirement. The first variable is based on the classification of people’s first job. In its original form in the questionnaire, this variable used the 11-item scale corresponding to the Swiss CSP (Levy and Joye 1994), later collapsed into six levels, to make them fit to the Goldthorpe categories (Bergman and Joye 2001; Goldthorpe et al. 1980). Accordingly, the modalities and corresponding distributions in the sample are: Managerial positions (13.6 %), intermedi- ary jobs (e.g. lawyers5) (10.5 %), self-employed (3.5 %), white-collar jobs (39.1 %), blue-collar jobs (31 %), and inactive, for those who reported that they did not work (2.4 %)—the latter category being almost exclusively 5 The typology of socio-professional categories which is used here is designed to reflect the Swiss social structure and can in certain regards be counter-intuitive. This particularly concerns the distinction between intermediary jobs and self-employed jobs, with both working independently. However, the main difference between the two is that intermediary professions have a higher degree of academic level, and thus includes such professions as lawyers, dentists or physio-therapists. Typical professions which would be categorized as self-employed professions are plumbers, chauffeurs or shopkeepers. 181 made up of women. Then, we have the same measure but referring to the last job. Compared to the first measure, people moved away from blue-collar professions (only 16 %), and a growing share were inactive (up to 7 %), as well as self-employed (9.4 %), workers in managerial jobs (a strong increase to 25.6 %), or in intermediary jobs (11.5 %). White-collar jobs decreased for the end of respondent’s work life to 30.4 percent. Then, we use three binary indicators relating to the more recent work-life. We start with whether people reported as still having been working occasion- ally or regularly in a paid job at the time of the first interview in 2011. Then, using the same information from the second wave of the survey in 2016, we constructed two measures capturing the temporal evolution: Whether people have given up or started a job. In the VLV panel, 8.6 percent have stopped a job between 2011 and 2016, 4.3 percent reported having taken up a job and a mere 2.5 percent continued to work between the two surveys, either regularly or occasionally. 8.3.4 Confounding and Stratification Variables Given that we study the elderly population, we have to account for a number of specific dynamics and circumstances, which can strongly influence the financial situation and whether respondents still work or not. Firstly, the most central dimension that we have to account for is age, since at higher ages the ability to work decreases, consumption patterns change, and potentially the need for care services increases. Next, we add sex as a confounding variable in all models to detect gender-specific patterns, which are often a key aspect of both ageing and poverty. Furthermore, we have to distinguish whether people live at home (95.7 %) or whether they live in a care institution (4.3 %) and finally, we also take into consideration differences in health among the respondents since health could be a financial burden, particularly in old age. We do so by using the self-rated health scale. While this scale does have a strongly subjective dimension, it is generally considered a good and simple measure for people’s physical condition (DeSalvo et al. 2006; Idler et al. 1999). 8.3.5 Modelling Approach Our analyses can be divided into a first block focusing on the cross-sectional VLV1 data with the aim of determining the link between working at the age of retirement and social stratification. In this first block, we distinguish between the two poverty measures: income-poverty and subjective poverty. For each, we proceed in an identical manner by constructing nested regression models, stepwise adding covariates starting with single covariates to assess their 182 unique effect, then proceeding to including two at the same time to check for mediation or moderation effects. More specifically, we start with a basic model that only shows the baseline effects for the confounding variables. Then we move to a model where we assess the effect of people’s first job (model 2). Following this, we do the same with people’s last job (model 3). Next, we assess the unique effect of still working at the age of retirement (model 4). Then, we include people’s first job as a measure of social stratification as well as whether they are still working (model 5). The last model then does the same with people’s first job and the work activity measure (model 6). For the second block of analyses, we shift our focus to the temporal dynam- ics of poverty where our target variables are the four types of trajectories that have been described in section 8.3.2. For each trajectory type, we proceed in an identical way as for the static analyses described above: We start out with a confounder model (model 1), then move on to people’s first job (model 2), estimate another model with people’s second job (model 3), then stepwise add the three possible scenarios for the respondent’s situation with regards to working: whether they have worked at both time-points (model 4), whether they have taken up a job (model 5) or have given up a job (model 6). If any of these event-measures turns out statistically significant, we will then proceed to building additional nested models to test for mediation/moderation effects with the measures of social stratification. As in the first block, these analyses are performed separately for each of the poverty measures. 8.4 Results 8.4.1 The role of social stratification and work after the age of retirement for poverty in old age Table 1 shows the results of the six binomial logit regression models for factors affecting income poverty. The confounder model (m1) reflects findings that were expected based on research in Switzerland and internationally: Above all, it shows that income poverty in old age clearly affects women more than men, as shown by an odds ratio of 1.31. Furthermore, self-rated health negatively affects poverty (OR 0.85), meaning the better health respondents indicate, the less there is a risk for poverty. While this result has been found elsewhere, the exact underlying mechanism appears unclear (McDonough et al. 2005). Living in a care-home does not seem to be significantly affecting OAP. In model 2, a strong link can be found between the type of job people have carried out at the beginning of their lives and the risk of poverty in old age, many years later. Managerial jobs (OR 0.39) and intermediary profes- 183 sions (OR 0.69) are protective when compared to the reference category of white-collar workers, while starting as a blue-collar worker clearly increases the risk (OR 1.87). For the same indicator at the end of people’s professional life (model 3), the patterns remain similar with only marginal shifts. From a more global perspective, this means that the upper and lower ends of the social structure almost directly relate to the respective risks of poverty in old age, which represents an almost perfect illustration of the social stratification paradigm. Another interesting result concerns self-employed workers who are also found to be highly likely to be in a situation of OAP (OR 3.84). Next, one of our main interests in this chapter, the relationship of still working at the age of retirement reveals as something which concerns people with more favourable financial means: the odds ratio of 0.59 shows that being professionally active beyond the legal retirement age actually signifies less risk for OAP than not working. This finding stands in complete opposition to our hypothesis according to which we argued that working at the age of retirement might be positively associated with OAP, since retired citizens with modest financial needs might be pushed to work to earn an income in order to afford daily expenses. Model 5 and 6 then tested for mediation Table 1: Poverty in old age in Switzerland 2011—regression on income below the absolute poverty line Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Age 1.01 1.00 1.01 1.00 1.00 1.00 Women 1.31*** 1.42*** 1.24* 1.28** 1.41*** 1.21* Living in Care Home 1.01 1.02 0.98 1.03 1.04 0.99 Self Rated Health 0.85*** 0.87** 0.87** 0.84*** 0.87** 0.86** First Job Upper (Ref. White Collar) 0.39*** 0.41*** First Job Self-employed 3.84*** 3.98*** First Job Intermediate 0.69* 0.68* First Job Blue Collar 1.87*** 1.90*** First Job Inactive 2.19*** 2.20*** Last Job Upper (Ref. White Collar) 0.41*** 0.42*** Last Job Self-employed 2.28*** 2.41*** Last Job Intermediate 0.45*** 0.43*** Last Job Blue Collar 2.05*** 2.03*** Last Job Inactive 1.02 1.02 Still working 0.59** 0.68* 0.60** Constant 0.23*** 0.22*** 0.24** 0.34* 0.30** 0.36* Observations 2,500 2,483 2,460 2,480 2,463 2,440 Log Likelihood -1,295.00 -1,229.00 -1,196.00 -1,277.00 -1,212.00 -1,180.00 Akaike Inf. Crit. 2,600.00 2,478.00 2,413.00 2,566.00 2,447.00 2,382.00 Note: Odds ratio of binomial logit regression. Significance: * p < 0.1, ** p < 0.05, *** p < 0.01 Source: VLV survey, 2011 & 2016, own calculations 184 effects between working in old age and social stratification, but yielded no significant results, reflected by only marginal and thus negligible shifts in the baseline odds ratios shown in models 2 and 3. Table 2 displays the results of the regressions on the variable making ends meet, our subjective measure of poverty. Overall, we find almost identical results over all models as for income poverty. The confounder model shows a higher risk for women and seniors in poor physical health. Regarding people’s first and last job we also find strong coherence with social stratification theory, but mostly for lower social strata, namely blue-collar workers where the odds ratio of 1.69 for the variable first job and 1.89 for last job signifies almost a double risk of OAP compared with the white-collar workers. Working at the age of retirement then also turned out as being negatively associated with OAP but not showing any mediation or moderation effects when included at the same time in a single model. Table 2: Making ends meet in old age in Switzerland 2011—regression on having difficulties in making ends meet financially Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Age 0.98*** 0.98*** 0.98*** 0.97*** 0.97*** 0.98*** Women 1.32** 1.49*** 1.34** 1.28** 1.47*** 1.33** Living in Care Home 1.61 1.50 1.57 1.62 1.51 1.59 Self Rated Health 0.60*** 0.63*** 0.62*** 0.61*** 0.64*** 0.63*** First Job Upper (Ref. White Collar) 0.72 0.78 First Job Self-employed 0.86 0.89 First Job Intermediate 0.54** 0.57** First Job Blue Collar 1.69*** 1.73*** First Job Inactive 0.81 0.80 Last Job Upper (Ref. White Collar) 0.57*** 0.60*** Last Job Self-employed 1.42* 1.48** Last Job Intermediate 0.44*** 0.45*** Last Job Blue Collar 1.89*** 1.91*** Last Job Inactive 0.66 0.65* Still working 0.45*** 0.50*** 0.49*** Constant 4.87** 3.47* 4.05** 6.73*** 4.41** 5.26** Observations 2,907 2,881 2,854 2,884 2,858 2,831 Log Likelihood -1,138.00 -1,108.00 -1,077.00 -1,127.00 -1,098.00 -1,068.00 Akaike Inf. Crit. 2,285.00 2,235.00 2,173.00 2,266.00 2,219.00 2,157.00 Note: Odds ratio of binomial logit regression. Significance:* p < 0.1; ** p < 0.05; *** p < 0.01 Source: VLV survey, 2011 & 2016, own calculations 185 8.4.2 Adding the temporal dimension: the poverty trajectories 2011–2016 We now shift our attention to the temporal dimension. Here, the main finding is that poverty in old age does not appear completely static but that there are dynamic movements in and out of it. For income-poverty (Figure 1), only 8.7 percent appear to be poor in both waves of the survey. This number has to be contrasted with the estimate of 21.4 percent for people living in income-poverty in 2011. Hence, our data shows that less than half of those experiencing poverty at one time-point continuously remain in the same sit- uation and thus experience structural poverty. At the same time, there are few but still non-negligible dynamic movements into (12.8 %) and out of poverty (4.1 %). For subjective poverty (Figure 2) the results once again are similar: While the large majority (80.9 %) never indicated to have difficulties to make ends meet between 2011 and 2016, only 6.5 percent reported to feel so in both waves of the survey. Equally important (8 %) is the share of people that started to feel under a financial strain, and a marginal share (4.5 %) reported having stopped feeling like it was difficult to make ends meet. 8.4.3 The Role of Social Stratification and Work on Poverty Trajectories in Old Age In this section we explored the factors that influence the identified poverty trajectories. We once again performed this analysis for our two measures of poverty separately: income-based poverty and subjective poverty. Here, we must point out, that while we performed regression analyses for each of the four possible trajectories and for both poverty measures, we only display the results for two trajectories: ‘stable poverty’ and ‘falling into poverty’. The reason for omitting ‘stable non-poverty’ is that the results are perfectly analogous to those found for stable poverty. The trajectory for ‘moving out of poverty’ is also omitted because of insufficient statistical power given that only a small share of individuals experience this development. Table 3 shows the results for the two income-poverty trajectories ‘being stable in a situation of poverty’—what might be described as structural pov- erty in old age—and ‘falling into poverty’. Results once again provide strong support for a social stratification view on OAP and thus means this theo- retical framework does not only apply to people’s situation at a given point in time but also in a longitudinal perspective. For stable poverty, the effects for blue-collar workers, i.e. those at the lower end of the social structure, was most consistent with those from the static analyses. Accordingly, having started one’s professional life as a blue-collar worker increases the risk of 186 Figure 1: Frequency of income-poverty trajectories 2011–2016 Stable in poverty Moving out of poverty Falling into poverty Stable in financial security / not poor 0% 20% 40% 60% 80% Source: VLV survey, 2011 & 2016, own calculations Figure 2: Frequency of subjective poverty trajectories 2011–2016 Not had difficulties to make ends meet 2011/2016 6.5% Stopped having difficulties making ends meet 4.5% Started to have difficulties to make ends meet 8% Not had difficulties to make ends meet 2011/2016 80.9% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Source: VLV survey, 2011 & 2016, own calculations being in a situation of structural poverty by more than a factor 3 (OR 3.31), compared to white-collar workers. Comparing this with the same indicator but at the end of people’s professional careers, the effect even increases to an odds ratio of 3.8. Contrasting with the results for the static analyses, we find no effect for the upper end of the social structure which we aimed to capture 187 with the category of upper managerial professions. Furthermore, those who were self-employed also show strong associations (OR 7.7 for the first and 3.6 for the last job) with OAP. Again, this is consistent with the fact that this group is somewhat at a disadvantage in the Swiss pension system. They have been shown to have more modest financial means in old age and to rely more substantially on the basic public pension scheme (AHV/AVS) and on private savings (Gabriel et al. 2015). In a longitudinal perspective, work (not having worked continuously at the time of both surveys, having given up a job, having taken up a new one) does not seem to have any discernible effect on stable poverty. As has been previously mentioned, this result can also be attributed to the insufficient statistical power of these models. For the trajectory of falling into poverty, people whose first job was in a blue-collar profession face a significantly increased risk (OR 1.81) to fall into poverty while ageing. The same applies to people who indicated having been inactive (OR 5.55). At the same time, people whose last job was in a managerial position have a reduced risk of falling into poverty (OR 0.42) than white-collar workers. Regarding people’s last job, the only significant effect concerns managerial jobs which appears as a protective factor. These findings are once again in line with a social strati- fication view, meaning that it not only drives structural poverty but also the movements into poverty. Analogous to table 3, table 4 shows the results for the two trajectories of ‘stable poverty’ as well as ‘falling into poverty’ but, this time employing the subjective poverty measure, For stable subjective poverty, we find less evident patterns compared to the findings for the static analysis for 2011 as well as with the results for the income-based poverty trajectories. The only signifi- cant result can be found for intermediate professions. The latter are found to have a reduced risk to have difficulties making ends meet (OR 0.27). As this category includes well-paying jobs such as lawyers or dentists that can be attributed to the higher end of the social hierarchy, this can nevertheless be taken as evidence for our social stratification hypothesis. The absence of effects for social stratification variables might have to do with the fact that in the subjective evaluation of poverty, there are multiple psychological factors at play, such as long-term coping strategies—i.e. those who have always lived with modest financial means somehow get used to this situation, regardless of social background and previous profession (see Henke 2020). However, in the absence of additional psychological variables these effects are difficult to quantify. The absence of any effects for work might once again be explained on the grounds of lacking statistical power. 188Table 3: Poverty trajectories in old age in Switzerland 2011‒2016 – Regressions on poverty trajectories ‘Stable poor’ and ‘falling into poverty’ Poverty trajectory type Stable income-poor Falling into income-poverty Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Age 0.99 0.99 0.99 0.99 0.99 0.99 0.96** 0.96** 0.96** 0.96** 0.96** 0.97* Women 0.97 1.36 1.02 0.97 0.99 0.99 0.86 0.87 0.70 0.85 0.85 0.89 Living in Care Home 0.52 0.49 0.48 0.52 0.50 0.52 0.60 0.58 0.60 0.60 0.60 0.59 Self Rated Health 0.70** 0.68** 0.70** 0.70** 0.70** 0.69** 0.71*** 0.70*** 0.70*** 0.71*** 0.70*** 0.70*** First Job Upper (Ref. White Collar) 0.82 0.67 First Job Self-employed 7.70*** 1.67 First Job Intermediate 0.44 1.16 First Job Blue Collar 3.31*** 1.81** First Job Inactive 0.0000 5.55** Last Job Upper (Ref. White Collar) 0.53 0.42** Last Job Self-employed 3.60*** 1.27 Last Job Intermediate 0.33* 1.21 Last Job Blue Collar 3.80*** 1.53 Last Job Inactive 1.20 0.88 Stable working 0.91 0.52 Take up a job 1.58 0.94 Give up a job 1.34 1.57 Constant 0.67 0.57 0.60 0.68 0.58 0.55 7.34 11.30 11.20 8.69 7.48 5.33 Observations 902 895 888 902 902 902 880 873 866 880 880 880 Log Likelihood -241.00 -222.00 -217.00 -241.00 -240.00 -240.00 -307.00 -299.00 -294.00 -306.00 -307.00 -306.00 Akaike Inf. Crit. 491.00 464.00 455.00 493.00 492.00 493.00 623.00 618.00 608.00 624.00 625.00 623.00 Note: Odds ratio of binomial logit regression. Significance:* p < 0.1; ** p < 0.05; *** p < 0.01 Source: VLV survey, 2011 & 2016, own calculations 189 Table 4: Making ends meet trajectories in old age in Switzerland 2011‒2016 – Regressions on subjective poverty t rajectories ‘continued having difficulties’ and ‘started having difficulties’ 2011‒2016 Difficulties in making ends meet in both surveys Started having difficulties Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Age 0.97 0.97 0.97 0.97 0.97 0.97 0.98 0.98 0.98 0.98 0.98 0.98 Women 1.25 1.47 1.17 1.26 1.23 1.31 0.96 1.01 0.82 0.96 0.94 0.97 Living in Care Home 0.93 0.93 0.86 0.93 0.96 0.93 0.52 0.52 0.47 0.52 0.53 0.51 Self Rated Health 0.49*** 0.50*** 0.49*** 0.49*** 0.48*** 0.48*** 0.70** 0.72** 0.71** 0.71** 0.70** 0.70** First Job Upper (Ref. White Collar) 1.02 0.33** First Job Self-employed 0.87 0.57 First Job Intermediate 0.67 0.65 First Job Blue Collar 1.66 1.84** First Job Inactive 1.35 2.49 Last Job Upper (Ref. White Collar) 0.51 0.49* Last Job Self-employed 1.08 1.36 Last Job Intermediate 0.27** 0.51 Last Job Blue Collar 1.68 2.07** Last Job Inactive 0.78 1.61 Stable working 1.16 0.83 Take up a job 0.59 0.52 Give up a job 1.85 1.23 Constant 7.28 5.56 7.40 6.96 8.87 5.01 1.42 1.56 1.93 1.49 1.65 1.25 Observations 942 935 928 942 942 942 925 918 911 925 925 925 Log Likelihood -213.00 -208.00 -203.00 -213.00 -213.00 -212.00 -252.00 -241.00 -242.00 -252.00 -252.00 -252.00 Akaike Inf. Crit. 436.00 436.00 426.00 438.00 438.00 436.00 514.00 501.00 504.00 516.00 515.00 516.00 Note: Odds ratio of binomial logit regression. Significance:* p < 0.1; ** p < 0.05; *** p < 0.01 Source: VLV survey, 2011 & 2016, own calculations 190 As for the final trajectory, which contains individuals who have indicated as having started to have difficulties to make ends meet when they were inter- viewed for the second time in 2016, the results largely mirror the findings for the corresponding income poverty trajectory. For those who shifted towards having difficulties making ends meet, managerial professions (OR 0.33 for the first job and 0.49 for the last job) again seem protected, while blue-col- lar workers (OR 1.84 and 2.07, respectively) show a significantly increased risk. Overall, the results for subjective poverty are highly in line with the findings for income-poverty and confirm that social stratification is as much a dominant force for OAP at a specific given time, as it is for how poverty evolves over time. 8.5 Discussion and Conclusion In this chapter, we have set out with the overall aim to evaluate the role of work for old age poverty (OAP). As part of this undertaking, we developed three contributions to the existing literature. Firstly, while multiple studies have assessed the relationship between past work-lives (which closely relates to social stratification) and OAP, we extended this focus by including work after the age of retirement. Secondly, we explored the relatively understud- ied temporal dimension of OAP, notably with regards to what factors drive changes in OAP over time. Thirdly, we addressed a continuing shortcoming in existing research on OAP that is a narrow focus on income-poverty alone, by considering an additional poverty measure, based on a subjective evaluation of respondent’s financial situation. Our theoretical framework consisted of a social stratification view and critical gerontology perspective. We empirically tested our hypotheses using two waves of the VLV survey. For the results from the ‘static’ analyses in 2011, we found that for both poverty indicators work beyond the age of retirement turned out the contradict our initial hypothesis as it was inversely associated with OAP, meaning that it is individuals in stable financial conditions who continue to work. Naturally, it is also possible that the causality has to be interpreted in the other direc- tion, meaning that because people continue to work, they are less susceptible to OAP. What is undisputed, however, is that there are no interactions with the original pattern of social stratification, meaning that working in old age represents a distinct phenomenon. For the dynamic analyses, both income-poverty and subjective poverty revealed themselves as by no means structural. There were non-negligible movements into and out of it. Regarding the factors that influence the trajec- 191 tories, we found coherent findings relating to social stratification across the ‘stable’ trajectories but also for falling into poverty, whereas the most consis- tent findings concern the lower end of the social hierarchy with blue-collar workers found to be almost always at a disadvantage. The dynamic analyses also demonstrated a high degree of coherence across both poverty indicators. However, with regards to work after the age of retirement, this factor played no role anymore for poverty trajectories. But as has been argued throughout the results section, this might also be linked to lacking statistical power. Based on these findings, most of our hypotheses can be accepted. The most straightforward result concerns the coherence between the subjective and income-centred poverty measure. This raises the conclusion that the level of income does indeed signify a subjective reality of experiencing financial strain, and that this reality is strongly driven by socio-economic position. This goes against the argument presented by Guggisberg and Häni (2014) who claimed that income-poverty might not be problematic per se, because retired possessed financial assets that could be used to offset the lack of liquid income, or that the patterns of consumption naturally decrease due to ageing, so that the level of stress perceived by the concerned individuals was expected to be negligible. As a whole, the patterns we observed in this chapter strongly support the social stratification and critical gerontology view. Working beyond the age of retirement, however, returned unexpected results in more than one way: Not only did the negative association between still working and OAP go in the opposite direction than what we expected, but it also appeared as being independent of social stratification patterns. Before we move on to our hypothesis regarding the social stratification and the temporal dynamics of OAP, we would like to stress the importance of the result that OAP is not simply a structural phenomenon. In fact, our findings actually stand in stark contrast to the majority of the literature on OAP, which, explicitly or implicitly, conceptualizes OAP as something being mainly static. In fact, we have ourselves made this assumption in previous work, incorrectly as it turns out, based on the idea that incomes at the age of retirement almost exclusively stem from a form of pension and thus remain relatively stable across time. But while incomes might remain stable, the rest of people’s life does not. As is the case for the population overall, retired citizen’s living arrangements and financial expenditure can change due to a variety of reasons, including requiring medical care at home, moving into a retirement home, or losing a partner. This brings us to one of our main findings with regards to our theoretical framework: our results for poverty trajectories show that social stratification seems not only to have a dominant effect on poverty at a specific time point, 192 but also on how it evolves over time. As has been pointed out previously, this view is also in line with a critical gerontology perspective that emphasizes how patterns of socio-economic inequalities become even more pronounced in old age. The only result that goes against our initial hypothesis is that work after retirement is negatively associated with OAP, seems independent from social stratification and does not play a significant role in temporal dynamics. In summary, these results leave us, once again to conclude to what tremen- dous extent poverty in old age is influenced by social stratification dynamics. This has been observed before, but only for income-poverty and only using a cross-sectional approach, and now has been confirmed for subjective poverty as well as for the most important trajectories as well. This also leaves us, with the persisting conundrum of how social stratification affects OAP. As we have previously mentioned, the exact causal mechanism for this strong relationship is difficult to determine and is most likely the result of multiple forces. However, a contributing factor is the Swiss pension system, which has a certain degree of social stratification built into it: High earners are able to heavily contribute to the so-called second pillar (professional pension plan) and often have the means to have a third pillar (private pension plan). Blue-collar workers, in contrast, have much more modest second pillar pensions and are less likely to dispose of disposable income allowing them to contribute to a third pillar. For the self-employed, this effect seems even more straightforward: they have no compulsory second pillar and thus mainly have to rely on AVS rent (unless they managed to contribute voluntarily to a third pillar). At the same time, differences between socio-economic position might also arise through the so- called non-take up (NTU) phenomenon related to supplementary benefits, a welfare benefit that supplements particularly small first pillar rents: Those with a better level of education such as white-collar workers might be less at risk of NTU than their blue-collar counterparts. Finally, since in this chapter we tried looking back at the past in order to then look ahead and formulate how the future of work might affect poverty in old age, we believe two developments play a key role. The first is improved social security in old age and the transformation of the elderly population. Over the last three decades, living conditions have greatly improved for the elderly population in Switzerland, including a reduction in income-poverty (Gabriel 2015; Oris et al. Forthcoming)—despite a significant rise in life expectancy. Even though there are signs of persisting inequalities, the effect of the welfare state with the establishment of the 3-pillar pension was con- siderable. This might hint at a continuation of the past trends for the future: the continuing improvement of the system of social security and improving material conditions for the elderly. Also, Swiss residents have also gotten 193 much better educated over the observed period with a decrease of educational inequalities, a tendency which also becomes evident for the elderly population, leading to less people in the lower social categories, which in this study have shown to be vulnerable to OAP. However, there is a second important dimension to take into consider- ation: the transformation of life courses, the labour market and with that, the patterns of contribution to pension schemes. The cohorts that we have so far studied are strongly characterized by standardized, gendered life-courses and there is practically no evidence for de-standardization of life courses among them (Widmer and Ritschard 2009). The dynamics of de-standardization were strongly fueled by the transformation of the economy with a rise of flexible and precarious work conditions. The first generations that will have their work-biographies strongly characterized by these tendencies are those born after 1960 (Brückner and Mayer 2005). It can be expected that the rise of such new work conditions will negatively affect the aforementioned cohort’s capacity to fully contribute to the pension scheme, even though empirical proof of the relationship between work trajectories and OAP is still lacking (Gabriel et al. 2015; Oris et al. 2017). In fact, Ganjour and colleagues state that ‘the degree of inequalities is more determined by social characteristics of individuals than by their occupational trajectories’ (Ganjour et al. 2016, 711). Yet, if the aforementioned dynamics lead to more self-employment—shown in this study to increase the risk for OAP—and drastically fragmented work biographies, this might affect future pensions and OAP. With this opposition between the positive dynamics of continuing improve- ments among the elderly population, fueled by a continuing democratization of education and other factors, and the negative effects of de-standardization and a rise of more instability in western society, it is difficult to determine which force will dominate and we refrain from formulating any more speculations. What we do want to emphasize, however, is that the effect of de-standardized life courses and work trajectories has to be better explored and should also be taken into account when discussing the reform of social security systems in Switzerland—a suggestion that we share with Frommert (2013) who has addressed this question for the German context. Also, based on our insights in this chapter regarding the longitudinal dimension of OAP we strongly encourage additional research focusing on poverty trajectories, rather than simply a static classification based on a one-time observation, as well as the factors that influence them. In light of the ongoing discussions on the reform of the pension system, we believe it is crucial to continue to advance our understanding of OAP in order to prevent those who are most vulnerable in old age, from being even more at a disadvantage in the future. 194 8.6 References Arent, Stefan and Wolfgang Nagl. 2010. A Fragile Pillar: Statutory Pensions and the Risk of Old-Age Poverty in Germany. FinanzArchiv / Public Finance Analysis, 66(4): 419‒441. Bane, Mary Jo and David Ellwood. 1983. Slipping Into and Out of Poverty: The Dynamics of Spells. The Journal of Human Resources, 21(1): 1‒23. Bergman, Manfred Max and Dominique Joye. 2001. Comparing Social Strati- fication Schemas: CAMSIS, CSP-CH, Goldthorpe, ISCO-88, Treiman, and Wright. Cambridge Studies in Social Research, 10: 1‒35. Berthoud, Richard, Morten Blekesaune, and Ruth Hancock. 2009. Ageing, Income and Living Standards: Evidence from the British Household Panel Survey. Ageing & Society, 29(7): 1105‒1122. Bonoli, Giuliano. 2003. Two Worlds of Pension Reform in Western Europe. Comparative Politics, 25(4): 399‒416. Börsch-Supan, Axel. 2014. Ökonomie einer alternden Gesellschaft. Perspektiven der Wirtschaftspolitik, 15(1): 4‒23. Börsch-Supan, Axel, Agar Brugiavini, Hendrik Jürges, Arie Kapteyn, Johan Mack- enbach, Johannes Siegrist, and Guglielmo Weber. 2008. First Results from the Survey of Health, Ageing and Retirement in Europe (2004‒2007). Starting the Longitudinal Dimension. Mannheim: Mannheim Research Institute for the Economics of Aging (MEA). Brückner, Hannah and Karl Ulrich Mayer. 2005. De-Standardization of the Life Course: What it Might Mean? And if it Means Anything, Whether it Actually Took Place? Advances in Life Course Research, 9: 27‒53. Budowski, Monica and Christian Suter. 2002. Living Arrangement and the Dynamics of Poverty: Lone Parenting in the Life Courses of Mothers. Schwei- zerische Zeitschrift für Soziologie, 28(2): 319‒339. Budowski, Monica and Robin Tillmann. 2006. La pauvreté persistente : un phénomène de classe, de cumul de désavantage ou d’individualisation ? Swiss Journal of Sociology, 32(2): 329‒348. DeSalvo, Karen B., Nicole Bloser, Kristi Reynolds, Jiang He, and Paul Muntner. 2006. Mortality Prediction with a single General Self-Rated Health Question. Journal of General Internal Medicine, 21(3): 267‒275. Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton: Princeton University Press. Frommert, Dina. 2013. Erwerbsverläufe und Alterssicherung. Duisburg-Essen: Univ.-Bibl. 195 Gabriel, Rainer. 2015. Inequalities Within Progress: Social Stratification and the Life Course Among the Elderly Population in Switzerland 1979‒2011. PhD dissertation, Faculté des Sciences de la Société, Université de Genève, Switzerland. Gabriel, Rainer, Michel Oris, Matthias Studer, and Marie Baeriswyl. 2015. The Persistence of Social Stratification? A Life Course Perspective on Poverty in Old-Age in Switzerland. Swiss Journal of Sociology, 41(3): 465‒487. Ganjour, Olga, Jacques-Antoine Gauthier, and Jean-Marie Le Goff. 2016. ‘Gender Inequality Regarding Retirement Benefits in Switzerland.’ In Gilbert Ritschard (Ed.), Proceedings of the International Conference on Sequence Analysis and Related Methods. Lausanne: Université de Lausanne, 693–716. Gavrilov, Leonid A. and Patrick Heuveline. 2003. ‘Aging of Population.’ In Paul Demeny and Geoffrey McNicoll (Eds.), The Encyclopedia of Population. New York: Macmillan, 32‒37. Goldthorpe, John H., Catriona Llewellyn, and Clive Payne. 1980. Social Mobility and Class Structure in Modern Britain. Oxford: Clarendon Press Oxford. Grusky, David Bryan. 2001. ‘Social Stratification.’ In Neil J. Smelser and Paul B. Balte (Eds.), International Encyclopedia of the Social & Behavioral Sciences. Oxford: Pergamon. Guggisberg, Martina and Stephan Häni. 2014. Armut im Alter. Neuchâtel: Bundesamt für Statistik (BfS). Haitz, Natascha. 2015. Old-Age Poverty in OECD Countries and the Issue of Gender Pension Gaps. CESifo DICE Report, 13(2): 73‒75. Harvey, David. 2011. A Brief History of Neoliberalism. Oxford: Oxford University Press. Hauser, Richard and Udo Neumann. 1992. Armut in der Bundesrepublik Deutsch- land. In Stephan Leibfried and Wolfgang Voges (Eds.), Armut im modernen Wohlfahrtsstaat. Wiesbaden: VS Verlag für Sozialwissenschaften. Haveman, Robert H. 2001. ‘Poverty: Measurement and Analysis.’ In Neil J. Smelser and Paul B. Baltes (Eds.), International Encyclopedia of the Social & Behavioral Sciences. Oxford: Pergamon, 11917‒11924. Hayward, Mark D. and Zhenmei Zhang. 2001. ‘Demography of Aging.’ In Kenneth Ferraro and Linda George (Eds.), Handbook of Aging and the Social Sciences. Cambridge, MA: Academic Press, 69–85. Henke, Julia. 2020. Revisiting Economic Vulnerability in Old Age: Low Income and Subjective Experiences Among Swiss Pensioners. Life Course Research and Social Policies. New York: Springer International Publishing. Himmelreicher, Ralf K. and Dina Frommert. 2006. Gibt es Hinweise auf zu- nehmende Ungleichheit der Alterseinkünfte und zunehmende Altersarmut? Der Einfluss von Erwerbs-und Familienbiographien auf die Rentenhöhe in Deutschland. Vierteljahrshefte zur Wirtschaftsforschung, 75(1): 108‒130. 196 Holden, Karen C., Richard V. Burkhauser, and Daniel A. Myers. 1986. Income Transitions at Older Stages of Life: The Dynamics of Poverty. The Gerontologist, 26(3): 292‒297. Idler, Ellen L., Shawna V. Hudson, and Howard Leventhal. 1999. The Meanings of Self-Ratings of Health. Research on Aging, 21(3): 458‒476. Ihle, Andreas, Ulrike Rimmele, Michel Oris, Jürgen Maurer, and Matthias Kliegel. 2020. The Longitudinal Relationship of Perceived Stress Predicting Subsequent Decline in Executive Functioning in Old Age Is Attenuated in Individuals with Greater Cognitive Reserve. Gerontology, 66(1): 65‒73. Kaeser, Laure and Jonathan Zufferey. 2015. Confronting Active Ageing with Em- pirical Evidence: A Structural Equation Model Approach. The Case of Older Migrants Living in Switzerland. Swiss Journal of Sociology, 41(3): 437‒463. Kohli, Martin. 2007. The Institutionalization of the Life Course: Looking Back to Look Ahead. Research in Human Development, 4(3‒4): 253‒271. Kohli, Martin and Camila Arza. 2011. ‘The Political Economy of Pension Reform in Europe.’ In Kenneth Ferraro and Linda George (Eds.), Handbook of Aging and the Social Sciences. Cambridge, MA: Academic Press, 251–264. Kwan, Crystal and Christine A. Walsh. 2018. Old Age Poverty: A Scoping Review of the Literature. Cogent Social Sciences, 4(1): 1478‒1479. Lalive d’Epinay, Christian, Jean-François Bickel, Carole Maystre, and Nathalie Vollenwyder. 2000. Vieillesses au fil du temps. Une révolution tranquille. Lau- sanne: Réalités sociales. Laslett, Peter. 1987. The Emergence of the Third Age. Ageing & Society, 7(2): 133‒160. Layte, Richard and Christopher T. Whelan. 2003. Moving in and out of Poverty: The Impact of Welfare Regimes on Poverty Dynamics in the EU. European Societies, 5(2): 167‒191. Leisering, Lutz and Stephan Leibfried. 2001. Time and Poverty in Western Welfare States: United Germany in Perspective. Cambridge: Cambridge University Press. Lessenich, Stephan. 2015. Die Neuerfindung des Sozialen: Der Sozialstaat im flexiblen Kapitalismus. Bielefeld: Transcript Verlag. Leu, Robert E., Stefan Burri, and Tom Priester. 1997. Lebensqualität und Armut in der Schweiz. Bern: Haupt. Levy, Rene and Dominique Joye. 1994. What is Switzerland’s Stratification like: Classes, Prestige Gradation, Professional Categories? International Sociology, 9(3): 313‒335. Ludwig, Catherine, Stefano Cavalli, and Michel Oris. 2014. “Vivre/Leben/ Vivere”: An Interdisciplinary Survey Addressing Progress and Inequalities of Aging over the Past 30 years in Switzerland. Archives of Gerontology and Geriatrics, 59(2): 240‒248. 197 McDonough, Peggy, Amanda Sacker, and Richard D. Wiggins. 2005. Time on My Side? Life Course Trajectories of Poverty and Health. Social Science & Medicine, 61(8): 1795‒1808. Möhring, Katja. 2015. Employment Histories and Pension Incomes in Europe: A Multilevel Analysis of the Role of Institutional Factors. European Societies, 17(1): 3‒26. Neugarten, Bernice L. 1970. The Old and the Young in Modern Societies. The American Behavioral Scientist, 14(1): 13‒24. Nolan, Brian and Ive Marx. 2009. ‘Economic Inequality, Poverty, and Social Exclusion.’ In Brian Nolan and Ive Marx (Eds.), The Oxford Handbook of Economic Inequality. Oxford: Oxford University Press. O’Rand, Angela M. 2006. ‘Stratification and the Life Course: Life Course Capi- tal, Life Course Risks, and Social Inequality.’ In Robert H. Binstock, Linda K. George, Stephen J. Cutler, Jon Hendricks, and James H. Schulz (Eds.), Handbook of Aging and the Social Sciences. 6th ed. Burlington: Academic Press, 145–162. Oris, Michel, Marie Baeriswyl, Aline Duvoisin, Rainer Gabriel, Myriam Girardin, Julia Henke, Barbara Masotti, Aude Tholomier, and Eric D. Widmer. Forth- coming. ‘The Sociohistorical Construction of Social and Economic Reserves Across the Life Course, and Their Use in Old Age.’ In Dario Spini and Eric Widmer (Eds.), Vulnerability Throughout Adult Life. Dynamics of Stressors, Resources, and Reserves. London: Palgrave Macmillan. Oris, Michel, Rainer Gabriel, Gilbert Ritschard, and Matthias Kliegel. 2017. Long Lives and Old Age Poverty: Social Stratification and Life-Course Insti- tutionalization in Switzerland. Research in Human Development, 14(1): 68‒87. Oris, Michel, Eduardo Guichard, Marthe Nicolet, Rainer Gabriel, Aude Tholomier, Christophe Monnot, Delphine Fagot, and Dominique Joye. 2016. ‘Repre- sentation of Vulnerability and the Elderly. A Total Survey Error Perspective on the VLV Survey.’ In Michel Oris, Caroline Roberts, Dominique Joye, and Michèle Ernst Staehli (Eds.), Surveying Human Vulnerabilities Across the life Course. Life Course Research and Social Policies. New York: Springer, 27‒64. Paugam, Serge. 1991. La disqualification sociale : essai sur la nouvelle pauvreté. France : Presses universitaires de France. Phillipson, Chris. 1981. State Policies and the Elderly. Ageing & Society, 1(03): 427‒432. Phua, Voon Chin, James W. McNally, and Keong-Suk Park. 2007. Poverty Among Elderly Asian Americans in the Twenty-First Century. Journal of Poverty, 11(2): 73‒92. Pilgram, Amélie and Kurt Seifert. 2009. Leben mit wenig Spielraum. Zürich: Pro Senectute Schweiz. Price, Debora. 2006. The Poverty of Older People in the UK. Journal of Social Work Practice, 20(3): 251‒266. 198 Rowntree, Benjamin Seebohm. 1901. Poverty: A Study of Town Life. New York: Macmillan. Samuel, Lawrence R. 2017. Aging in America: A Cultural History. Philadelphia: University of Pennsylvania Press. Sandoval, Daniel Al., Mark R. Rank, and Thomas A. Hirschl. 2009. The Increasing Risk of Poverty Across the American Life Course. Demography, 46(4): 717‒737. Schulz, James H. and Allan Borowski. 2006. ‘Economic Security in Retirement: Reshaping the Public-Private Pension Mix.’ In Robert H. Binstock, Linda K. George, Stephen J. Cutler, Jon Hendricks, and James H. Schulz (Eds.), Handbook of Aging and the Social Sciences. 6th ed. Burlington: Academic Press, 360‒379. SKOS, and CSIAS. 2013. Armut und Armutsgrenze – Grundlagenpapier der SKOS. Bern: SKOS. Smeeding, Timothy M. and Susanna Sandstrom. 2005. Poverty and Income Maintenance in Old Age: A Cross-National View of Low Income Older Women. LIS Working Paper Series. 398. January, 2005, https://www.econstor. eu/bitstream/10419/95568/1/483131938.pdf (accessed March 11, 2021). Swiss Federal Statistical Office. 2020. Szenarien zur Bevölkerungsentwicklung der Schweiz und der Kantone 2020‒2050. Neuchâtel: Swiss Federal Statisti- cal Office (SFSO). 28 May, 2020, https://www.bfs.admin.ch/bfs/de/home/ aktuell/neue-veroeffentlichungen.assetdetail.12847549.html (accessed March, 11, 2021). Vandecasteele, Leen. 2010. Poverty Trajectories After Risky Life Course Events in Different European Welfare Regimes. European Societies, 12(2): 257‒278. Vandecasteele, Leen. 2011. Life Course Risks or Cumulative Disadvantage? The Structuring Effect of Social Stratification Determinants and Life Course Events on Poverty Transitions in Europe. European Sociological Review, 27(2): 246‒263. Vlachantoni, Athina. 2012. Financial Inequality and Gender in Older People. Maturitas, 72(2): 104‒107. Walker, Alan. 1981. Towards a Political Economy of Old Age. Ageing & Society, 1(01): 73‒94. Walker, Robert, Roger Lawson, and Peter Townsend. 1984. Responses to Poverty: Lessons from Europe. Fairleigh Dickinson University Press. Wanner, Philippe and Alexis Gabadinho. 2008. La situation économique des actifs et des retraités. Aspects de la sécurité sociale. Rapport de Recherche 2008, No 1. Berne: Office Fédéral des Assurances Sociales. 28 February, 2008. https:// www.newsd.admin.ch/newsd/message/attachments/11722.pdf (accessed April 9, 2021). Widmer, Eric D. and Gilbert Ritschard. 2009. The De-Standardization of the Life Course: Are Men and Women Equal? Advances in Life Course Research, 14(1‒2): 28‒39. Yang, Yunjeong. 2011. No Way Out but Working? Income Dynamics of Young Retirees in Korea. Ageing and Society, 31(2): 265‒287. 199 9 ‘Simple Jobs’ for Disqualified Workers: Employability at the Bottom of the Labour Market Eva Nadai and Anna Gonon 9.1 Introduction Employability is a key issue in discourses and policies addressing the social consequences of labour market transformation driven by rapid technological change and globalization. Knowledge and skills are commonly regarded as the core ingredients of employability. Therefore, those labelled as unskilled, because they lack formal occupational qualifications are disqualified as un- employable. At best they seem fit for ‘simple jobs’, which do not require any occupational qualification beyond a short initial training on the job. Yet, routine manual and cognitive jobs are less and less available, which puts low-skilled workers at high risks of unemployment and precarity (Bolli et al. 2015; Crettaz 2018; Murphy and Oesch 2018; Nathani et al. 2017). In parallel to the marked decrease of low-skilled jobs, the proportion of workers without post-compulsory education has fallen; in Switzerland, only 11 percent of the population are unqualified, with differences according to age, gender, and nationality (BFS 2020). To the degree that low-educated workers have become a small minority, the lack of qualifications constitutes a stigma in its own right. The unskilled are discredited as deviant from the norm and their abilities are systematically underestimated (Gesthuizen et al. 2011; Solga 2008). The debate on the decline of low-skilled jobs rests on the assumption of a correspondence between the quality of jobs and the quality of workers. Yet, critical sociologists maintain that supposedly simple jobs are actually more complex than meets the eye and that upskilling is taking place within the remaining low-skilled jobs: depending on industry and job higher technical and social competences are required such as digital literacy, numeracy, com- munication abilities, and the like (Abel et al. 2014, Bosch and Weinkopf 200 2011; Hilf et al. 2018; Holtgrewe 2015; Krenn et al. 2014; Nickson et al. 2017). Skills or qualifications are neither fixed characteristics of workers nor objective features of a job. Rather, such categories are effects of valorization in particular historical and local contexts. To say that a job or a worker is ‘un- skilled’ implies a judgment on the social and economic worth of the activity or the person. Which skills are deemed valuable or are recognized as occupa- tionally relevant qualifications in the first place is determined by economic and political institutions, power struggles and conflicts that reach beyond the immediate workplace. The construction of qualification is a multi-level process involving educational systems, labour market policies, business enterprises, migration regimes, and the transnational division of labour in a globalized world (Atzmüller et al. 2015; Demazière and Marchal 2018; Hilf et al. 2018; Holtgrewe 2015; Hürtgen 2015; Krenn 2015; T ranchant 2018; Vallas 1990). This chapter focuses on the firm as a crucial site of valorization. Drawing on the theoretical framework of the Economics and Sociology of Conventions (EC/SC) it examines the relation between low-skilled jobs and workers from the perspective of quality conventions for labour. On what grounds and in which jobs are workers without formal qualifications still regarded as valuable labour and how is valorization tied to a firm’s coordination of production and its employment strategies? EC accords the enterprise a central role as ‘the principal actor deciding on the value of workers’ (Larquier 2016, 36). Hence, skills are not valuable in and of themselves but only inasmuch they are valued by a specific firm. We will show, first, that the value of low-skilled workers is contextual insofar it is just as much based on their fit into an existing work organization and company culture as on their individual abilities. Second, the relation between the fit into the coordination of production and value can go both ways: firms can select workers who fit the skills requirements of a job, or they can adapt technical and organizational formats to the abilities of the available workforce. Third, certain qualities are valued and devalued at the same time: they constitute significant criteria for the selection and evaluation of workers, yet they are not acknowledged as occupational qualifications. The analysis is based on data from an ongoing qualitative research on the constitution of the employability of low-skilled workers in five different industries.1 The following section presents the theoretical frame and describes the database of this study. In the third section we analyse the ‘simple’ jobs in the study firms and how these companies legitimize the use of low-skilled workers. The fourth section examines the qualities of workers that employ- ers expect: which abilities, behaviour, and personal traits characterize the 1 The research is funded by the Swiss National Science Foundation (grant 1000_175842). Together with the authors Robin Hübscher and Anna John collaborate on the project. 201 valuable worker in low-end jobs? The chapter concludes with a discussion of our findings. 9.2 Theoretical Framework and Empirical Database In everyday language the term qualification refers to the certified proof of a specific education and training. For the Economics and Sociology of Con- vention it has the more encompassing meaning of attributing properties to social objects as well as conferring a quality or value (Diaz-Bone 2018, 143). A prominent explanation for the low value of unqualified workers holds that educational attainment serves as an indicator of productivity and trainability; the lack of qualification, therefore signals low productivity (Gesthuizen et al. 2011). In contrast, Salais (2001) argues that labour and its value do not yet exist at the moment of selling it in the market, i.e. when employers and workers enter into an employment contract. What matters is not the work per se but ‘the (very material and singular) product of work’ (Salais 2001, 7). Hence, the value of labour is only realized at the moment of selling the product in the market. Hiring thus requires prognostic judgments on how the prospective worker contributes to production and market success. Valorization comprises two analytically distinct operations: valuation as ‘adding value to something’ and evaluation as ‘producing a judgment by as- sessing the value of something’ (Bessy and Chauvin 2013, 85). In their studies on recruitment and performance assessments EC/SC scholars demonstrate that these two operations are inseparable (Bessy et al. 1999; Bourguignon and Chiapello 2005; Eymard-Duvernay and Marchal 1997; Kozica and Brandl 2015; Larquier 2016; Marchal 2015; Marchal and Rieucau 2010). The main thesis is that the quality of (prospective) workers does not exist independently of the interlinked evaluation practices of different groups of actors, with the firm being the most important actor. In Eymard-Duvernay’s (2012, 11) concise formulation: ‘The judgment is the operation that creates the value.’ The dispositives of evaluation define and select the information that constitutes the basis for testing qualities like competence, productivi- ty, and performance. Attributions of quality refer to the alignment of the competencies and characteristics of the person with the specific exigencies of production (Diaz-Bone 2009, 180)—the value of workers is therefore contextual. In the labour market, there are at least two ‘critical moments’ (Boltanski and Thévenot 2011, 55) when the value of a person is at issue: the moment of recruitment and the (recurring) moments of performance evaluation during employment. 202 The notion of quality convention reflects the contextuality of valuation. It is a central premise of EC/SC that social situations are characterized by a plurality of possible logics of coordination and evaluation to which actors can refer to justify actions and evaluations. Quality conventions rest on overarching moral ‘orders of worth’ and provide evaluation frames and tests for situationally appropriate judgments on the value of persons, objects, and actions (Boltanski and Thévenot 2006). Conventions function as equivalence principles that allow the ranking of objects according to their relative worth. Of the eight conventions currently established in EC/SC (Diaz-Bone 2018, 146‒164), the market and the industrial convention are most relevant for the valorization of low-skilled workers. In the market convention value is determined by demand and tested in the situation of a transaction. Of little worth are those objects or persons that are not in demand and meet with rejection, for instance, workers whose productivity seems questionable because of some stigma (Nadai et al. 2019, 16). The industrial convention values predictability, efficiency, and professional expertise. The important evaluation criterion is the functionality of persons for the coordination of production: their productive capacities and their ability to ‘integrate themselves into the machinery, the cogwheels of an organization’ (Boltanski and Thévenot 2006, 205, original emphasis). Moreover, we argue that the firm is not just a site of production but also a social entity—a kind of community. Hence, the domestic convention too comes into play. In this convention value is tied to relations: the person is assessed within networks of interdependencies, trust, and loyalties. Trustworthiness, ‘membership in a closed universe’ (ibid. 166), and loyalty to superiors are prerequisites for the worth of the subordinate. Convention-based valuation is not just a discursive operation but must refer to external arrangements of objects for the assessment of the qualities and relative worth of social entities (Diaz-Bone 2018, 166). As mentioned before, the coordination of production is the foremost reference for the valuation of labour; formal qualification being the other main object for the assessment of individual workers. For EC/SC, both work organization and educational diploma are examples of ‘investments in forms’ that stabilize the practices of coordination and evaluation (Thévenot 1984; Thévenot 2001). The empirical study that provides the data for this analysis examines the constitution of employability as a pragmatic, dynamic, and interactive process (Promberger et al. 2008). It combines the perspective of unskilled workers, firms, and private and public labour market intermediaries using semi-structured qualitative interviews, which are complemented by documents. The focus is on five industries with a high percentage of low-skilled jobs and workers: catering trade, construction, cleaning, manufacturing, and retailing. 203 Firms were sampled from these industries, while the sample of workers covers a wider range of industries because of horizontal mobility in the course of employment trajectories. Intermediaries include commercial personnel staffing agencies and public agencies for the job placement of unemployed people. The database comprises thirty-three interviews with HR staff and managers of twenty-seven firms, ten interviews with intermediaries, and interviews with thirty-nine workers.2 The analysis presented here is based on the interviews with firms. Seven of the twenty-seven firms are SMEs with less than 250 employees. About two thirds of the interviewees are local HR managers, di- rect supervisors, production managers, or SMEs’ owner-managers who have operational responsibilities for the recruitment and performance evaluation regarding low-skilled workers. One third of the interviewees are specialized or high-level HR managers or branch managers who are not directly involved in recruitment and performance evaluation but have a profound knowledge of the firm’s employment strategies. The interviews pertained to the function of low-skilled workers for the coordination of production, employment con- ditions, work organization, recruitment practices (including the relations to temporary staffing agencies), and skills requirements. The study follows the principles of grounded theory methodology with respect to the intertwining of sampling, data collection and data analysis; data analysis is based on the coding techniques of Strauss and Corbin (Strübing 2014). 9.3 ‘Simple Jobs’? The Use of Low Skilled Workers in the Coordination of Production Asked what jobs they offer to low-skilled workers, managers often begin with a description of the lowest functional levels in their company, as the following statement by an HR manager in the retail industry illustrates: Regarding competences, it is really an assisting function. There is a boss telling you what to do. You are guided and supervised by someone experienced. And tasks are precisely defined within the context of the organization. The concept of a function as it is used in this statement refers to the abstract task that is defined independently of the person who performs it (Acker 1990, 148). Companies invest in the specification of tasks in order to stabilise a 2 To capture the dynamic nature of employability workers are interviewed up to three times over a period of three years. The number of interviews refers to the first interview wave. 204 certain coordination of production (Thévenot 1984, 16). As our data show, low-skilled workers perform a range of tasks with quite variable requirements. In catering, low-skilled workers primarily work backstage as ‘kitchen helps’: they clear tables, put dishes in dishwashers, store clean dishes, and they assist the trained cooks in preparing ingredients and arranging food on plates. They mostly perform manual routine activities. The work of preparing dishes is divided into highly specialized and hierarchic functions. ‘The chef reads out the order and then plates are prepared, and everybody has his own part, which he has to put on the plate’, an HR manager tells us. Before being served, each plate is checked by the chef. However, low-skilled employees work as waiters, too. This job implies more discretion and a broader variety of tasks, such as preparing one’s serving station (bread, cutlery, napkins) depending on the expected number of guests, taking orders, and communicating with customers. It also includes cognitive requirements, as waitresses need to have some knowledge about the menu, ingredients and allergens, and they must add bills, calculate change, or handle electronic cash registers. In retailing, the jobs performed by low-skilled workers are diverse, too. In large stores, they are mostly employed for routine tasks, such as stocking shelves or operating the till. In smaller stores, low-skilled workers are respon- sible for almost anything related to running the store: cleaning and tidying, taking delivery of the goods, preparing stock, presenting goods on the shop floor, serving customers and explaining to them the products. They perform thus not only what Pettinger (2004) calls routine service but also the more complex personal service. Besides manual and interactive activities, the job includes administrative tasks, such as ordering goods, settling daily takings, and depositing them in the bank at the end of the day. In the construction sector, low-skilled workers start in the function of a ‘handyman’. Since heavy physical labour performed with ‘shovel and pick’, has mostly been replaced by machines, as an HR manager explains, they mainly ‘clean up and carry stuff’. Autonomous decision-making and cognitive requirements are rather restricted in this job. However, it is common that after some years of experience or after having attended a course, workers are charged with more demanding tasks, such as formwork or operating machines. It is not unusual that low-skilled workers move up to the position of a general foreman. In industrial production, low-skilled workers are mostly employed in manual routine jobs. They pack goods, wash, and prepare food at the assembly line, and they are charged with re-filling material in machines. Task discretion and overall cognitive requirements are limited, although some companies have implemented teamwork and digital devices as means of work coordination 205 or charge low-skilled workers with quality control. Moreover, in most of the industrial companies of our sample, some low-skilled employees work in the higher-level function of machine operator. In the cleaning sector, the range of activities of low-skilled workers is very narrow. Most of them are employed in so-called ‘maintenance cleaning’. Ac- cording to the widespread ‘district system’ (Hieming et al. 2005, 115) workers always clean the same rooms following a standardized plan. Their job is thus very repetitive and does not imply any discretion. A regional manager of a large cleaning company describes the activities performed by a maintenance cleaner in detail: She prepares her cart. She checks if she has the right cloths. We work with four colours. The red cloth and the red detergent are for toilets. The yellow detergent and cloth are for mirrors and glass. Blue is for surfaces and green for kitchen areas […] Then she has a cleaning plan. It is a very simple plan. It has, activity, for example, vacuum- cleaning floor, cleaning table, cleaning cornices, maybe removing cobwebs, and then it has Monday until Friday, and then you have a cross for Monday and Wednesday and so she sees for every day what she must do […] When she is done, she goes down to the cleaning room, she checks if she has enough detergent, cleans her equipment, disassembles the vacuum cleaner and so on. And then she signs that she has performed her tasks today according to the plan. Compared to shop assistants or waiters, the contact with customers is very limited in cleaning, unless workers are promoted to the position of an ‘object manager’. In this function, they supervise a team of cleaners and they are responsible for contacts with customers (Benelli 2011). Explaining why they hire low-skilled workers, the interviewed managers refer to two main arguments. First, they argue that some of these jobs are ‘auxiliary’ tasks, i.e. jobs that were created to assist other functions. The job of a handyman on a construction site, for example, consists in being at the foreman’s disposition. Kitchen helps relieve the skilled cooks from the lengthy and painstaking task of cutting large amounts of vegetables and retail workers who replenish shelves and operate tills enable the self-service of customers (Pettinger 2006). The classification of these tasks as auxiliary mirrors the low valuation of the included activities. ‘Auxiliary’ jobs contain tasks that rank low regarding social prestige. Many of them involve cleaning activities. Moreover, they are considered as simple because they mostly consist of manual routine tasks that can be performed without technical knowledge. At least, the required skills have not been formalized—there is no corresponding 206 training in the vocational education system. Based on this evaluation, wages in these jobs are low. The formatting of auxiliary tasks as distinct jobs is typical of a work orga- nization which is based on a strong division of labour and hierarchy. Given the lower labour costs for low-skilled workers, it can be part of a cost-saving production model and employment strategy. The HR manager of a large catering business explains: When we look for a kitchen help or something, then there is actually almost nobody with training, but as a matter of fact, we do not look for anyone with training. First, because of the wage, and second, we made the experience that nobody with training wants to work as kitchen help. Referring to the labour supply and the skill level of the available workforce, on the one hand, the HR manager makes clear that on the other hand, the lower pay is the central reason for hiring low-skilled workers in this job. This employment strategy is based on the format of work coordination: instead of integrating the tasks of a kitchen help into the job profiles of a skilled cook or waitress, they are constituted as distinct jobs. This implies a strong operational division of labour. In manufacturing, the lower labour costs for low-skilled workers can be a decisive factor in the design of the production process, especially regarding the degree of automation. While there are still tasks that cannot be automated without producing too much waste, like for example the filleting of oranges and limes, in other cases, the labour of low- skilled workers is simply cheaper compared to the alternative of automation: People standing at the machine, putting toothbrushes in boxes, people who fill the machine with toothbrushes, you could automate everything, that’s no question. But depending on the batch size3, it isn’t profitable. And because of that, we will always need people doing manual work. (HR manager of consumer goods manufacturer) Since machines must be rebuilt depending on orders, automation only pays off for large quantities (Abel et al. 2014, 73, 190‒198). Work coordination is thus not determined by the available technologies, but also by employment strategies which in turn are influenced by the availability of cheap labour (Wickham 2011, 236). By formatting auxiliary tasks into distinct jobs instead of integrating them in the job profiles of skilled workers or automating them, 3 The batch size is the amount produced at a time without stopping the production process. 207 the employment of low-skilled workers can thus be an integral part of the companies’ production models. The second argument managers bring forward in order to explain the employment of low-skilled workers is the shortage of skilled workers. In this case, the employment of untrained workers is not embedded in a specific production model, but rather appears as a makeshift solution given the lack of trained applicants for jobs where apprenticeships exist, such as waitresses, sales assistants, or masons. The lack of skilled workers on the labour market is reported in several industries. The skills shortage is often explained by the decreasing attractiveness of an apprenticeship, for example of manual work compared to white-collar work: ‘Many just want to be in the office and not outdoors on the construction site anymore,’ an HR manager of a construction company says. The devaluation of some occupations in the service sector has been driven by the extension of opening hours, the increasing share of part-time jobs, and declining wages (Bosch and Weinkopf 2011, 182; Cian- feroni 2019). Accordingly, the HR manager of a retail company explained that ‘finding qualified workers for a part-time job with training is a lot more difficult.’ The devaluation of apprenticeships and occupations thus creates job opportunities for low-skilled workers. Accepting low-skilled job applicants for positions where formal degrees could be required is a common practice in industries with persisting skills shortage and can thus become part of a company’s employment strategy. The more common this practice, the more blurred the boundaries between low-skilled and skilled jobs become. Especially in industrial production and the construction sector, firms use their low-skilled employees in entry-level jobs as a recruitment pool for skilled positions. The production manager of a plastic manufacturing company explains: The market is not huge. We have employees in packaging that we notice, they are good. We might promote them a little. We’ve taken some of them in the machine hall and started an internal training. And we’ve trained them as machine operators. And now they work as machine operators. As this statement illustrates, there exist internal labour markets for low-skilled workers too. While in general, companies’ support of further training courses is particularly low among low-skilled workers (Wotschack and Solga 2014; Wotschack 2017), the companies in our sample selectively invest in the train- ing of low-skilled workers to meet their needs of skilled labour. Recruiting internal employees from entry-level jobs for skilled and supervising positions 208 also allows to use their implicit operational knowledge and to benefit from the authority that some long-standing workers enjoy among their colleagues. The use of low-skilled workers in skilled positions raises questions about the recognition of practical skills and experience. In the construction sector, the recognition of practical experience is institutionalized in the form of a nationally binding collective bargaining agreement, granting workers with a certain level of work experience the right to a higher wage category. Moreover, the industrial relations organizations have been investing in an extensive edu- cation programme for workers without vocational training in order to address the increasing skills shortage. In the absence of such formats that guarantee the valorization and development of practical skills, the employment of low- skilled workers can also reinforce the process of devaluation of professional skills in an occupation. In the retail industry, the increasing presence of low- skilled workers contributes to the process of dequalification (Demazière and Marchal 2018, 23). In our data, their employment is often commented with the statement that non-professional knowledge and competences count at least as much as professional skills (Hilf et al. 2018, also see 4). Having become part of a firm’s employment strategy, the systematic use of low-skilled labour can entail specific investments in the forms of work orga- nization. Several form investments in our data aim at the simplification of communication, because a large part of the low-skilled workers are migrants with limited command of the local language.4 A manufacturing company, for example, installed touch screens at the assembly line that show visual symbols such as laughing or crying smileys and numbers in the colours green, orange or red in order to inform workers about the accuracy of working steps and about the status of the whole production process. For instance, if a symbol is shown in red, it means that a machine is not running properly anymore and that the shift supervisor must be notified. Colours as means of codified communication are also used in cleaning companies, for instance, the standardized colours of cleaning cloths mentioned above or the ‘parrot plan’ to instruct workers about the work schedule. The rooms and areas to be cleaned on a given day of the week are marked with colour schemes that can be understood without reading skills in the local language. Other companies also use linguistically homogenous teams who can speak their own languages as a principle to overcome difficulties in communication. 4 In Switzerland, they count as ‘unskilled’ although they might have a formal educa- tion certificate from their home country. In many cases, foreign certificates are not recognized as equal to Swiss qualifications. 209 9.4 Low-Skilled and (de)Valued The jobs for low-skilled workers described by the firms of our sample cover a range from shovelling earth on a construction site to explaining to cus- tomers the origins of the food on their plates in a restaurant or calculating personalized discounts in a department store. Employers tend to ignore these differences with the sweeping statement that ‘everybody could do this’. If a job is so simple that it is within everybody’s reach, then those doing the job are interchangeable. Consequently, it would not matter who is doing those ‘unskilled’ jobs and firms might as well draw lots for them. This is obviously not what happens. If probed a little further, employers do elaborate on the skills needed for these ‘simple’ jobs and on the individual differences between basically interchangeable workers. Firms therefore still have to specify worker quality, albeit below the level of formal qualification, and they must select individuals possessing the required qualities out of an undistinguished mass of low-skilled workers (Demazière and Marchal 2018). Beyond the moment of selecting workers, qualities are also tested recurrently through informal observation on the job and formal performance assessments. In the recruitment process assessment is often implicit: managers refer to their ‘gut feeling’ to explain that a candidate is just right for the job (Hassler et al. 2019; Imdorf 2010). Nevertheless, this gut feeling is based on the observation of a worker’s actual performance during a trial period of varying duration before employ- ment, which is a common practice in low-skilled jobs (Hassler et al. 2019; Hieming et al. 2005; Marchal and Rieucau 2010). Moreover, some firms of our sample use standardized checklists for the assessment of the trial day, by which the recruiters’ intuition is controlled to a certain degree. Likewise, per- formance assessment during employment is a blend of the day-to-day informal observation by supervisors and yearly formal employee assessments, where the responsible supervisor typically checks, ‘has he performed quantitatively, performed qualitatively and are our values there’, as the HR manager of a large catering firm describes the logic of the evaluation.5 In public and scientific discourses on changing demands in the labour market, non-formalized skills are discussed under the label of key competen- cies (e.g. OECD 2019). Key competencies are supposed to be necessary and useful in any occupation, but especially important for those who lack formal knowledge and training. At the bottom of the labour market, the idea of key competencies is toned down to ‘basic competencies in the workplace’. Basic competencies comprise the same areas as key competency frameworks, but 5 A few firms of our sample do not use formal performance appraisals in low-skilled entry-level functions. 210 on a lower level: oral and written communication, numeracy, digital literacy, and various social competencies enabling teamwork and the like. The lan- guage of key or basic competencies does not appear in our interviews, but to some degree, it is recognisable in the description of the valuable worker. Yet, managers and supervisors emphasize different skills and expect much lower competence levels than political and educational discourses would have it. While the qualities of a ‘good’ worker in low-skilled jobs vary with industry, job, and work organization two features stand out in all five industries selected for our study: fitness and fitting in. Physical fitness is hardly ever mentioned in discourses of technological change and labour market transformation. Rather, discussions about digitali- zation or automation are mostly ‘disembodied’: they focus on the replacement of manual jobs by technology and on the increasing level of cognitive skills that workers need (Aepli et al. 2017; Nathani et al. 2017; OECD 2016). However, in the remaining jobs at the bottom of the labour market fitness in terms of strength, stamina, and overall physical health are still crucial, because workers lift heavy weights, they stand and walk for hours, they work night shifts, etc. (Hassler et al. 2019; Hieming et al. 2005; Krenn et al. 2014; Tranchant 2018). Both, the fit body and fitting in socially, take on contextually different forms. In manual jobs in construction, manufacturing, and cleaning, physical fitness comes in the shape of sheer bodily strength and robustness, or as dexterity. Robustness is assessed by an informal inspection of the body: recruiters look at the hands, posture, build etc. to estimate the worker’s suitability for the job: Well we don’t say show us 20 push-ups ((laughs)). But sometimes you see people and you’d think they’ll never make it […] We wouldn’t take on a woman, who’s 165 [cm] […] and rail-thin to hang up salami. For this job we do need a man. (HR manager, meat factory) In addition to physical strength, the good worker also possesses a certain mental hardiness. The ideal construction worker, for example, must be able to endure changing weather conditions and the bodily pain of a long workday’s exertion. ‘You must be a robust person; you can’t ache all the time,’ a manager stated.6 Apart from the informal impression that a worker is physically suited for the job, firms are legally obliged to arrange (recurrent) formal medical 6 Robustness and strength are gendered insofar employers have a male or female body in mind for jobs that are themselves gendered. However, the physical demands do not stem from the job per se. Rather, Sardadvar et al. (2015) argue, the gender-typing of the job influences whether it is perceived as physically demanding, hence requiring the quality of robustness. Dexterity is likewise gendered, namely as a female quality. 211 tests for certain jobs, e.g. if the work organization involves regular nightshifts. While physical fitness and health are obvious aspects of worker quality in manual jobs, it is an issue in service jobs too: in catering and retailing workers also lift goods, they unpack and process stock, clean shops and tables, are on their feet all day, and even though they do not work nightshifts, the irregular working hours are mentioned as an additional stress that worker must be able to cope with (Hieming et al. 2006; Monchâtre 2018; Pettinger 2006). For tasks that cannot be automated in a profitable way physical fitness is thus an indispensable quality: the worker’s well-functioning body takes the place of the machine to ensure the smooth coordination of production. Fitting in pertains to social skills, but not so much to the isolated, learnable skills listed in catalogues of key competencies. Rather, fitting in is a matter of attitude: the willingness and ability to adapt to the demands of the firm and the team. Essentially it means to know one’s place within the social order of the firm, which includes docility (Tranchant 2018, 126) and accepting the given employment conditions (Smith and Neuwirth 2009; Waldinger and Lichter 2003). Workers should show a well-balanced blend of adaptation and self-determination, as the HR manager of a restaurant chain said: ‘We don’t want someone who just does what he’s told but who thinks independently […] but a lot is already given.’ Fitting in encompasses general virtues like punctuality, reliability, precision, and flexibility regarding working hours, shifts, and tasks on the one hand; on the other hand, contextually specific requirements. On the most basic level, in the absence of specific qualifications, the willingness to work hard is the foremost distinguishing characteristic of the valuable low-skilled worker in all the industries of our sample. This attitude expresses itself in the often-used phrase of ‘seeing the work’, i.e. in carrying out necessary tasks without and beyond explicit orders. As Kleemann et al. (2019, 182) note, even in middle- and low-skilled jobs subjectivation is used as a functional resource to increase productivity. The worker who ‘is proud to be kitchen help at [the firm]’, a restaurant manager explains, ‘goes the extra mile, is happier, healthier, less absent from work, and, yeah, more productive.’ In other words, it is the productivity-enhancing subjective identification with the job and the particular company, the enthusiasm and loyalty of the happy kitchen help, that constitutes the value of the low-skilled worker. Fitting in through subjectivation is even more important in interactive service jobs: especially salespersons are expected to actively project their identification with the firm and the products they are selling. For the manager of a low-end retail chain the lack of formal qualification does not matter if the salesperson feels and demonstrates ‘passion’ and regards the firm as ‘an extremely cool brand’. When firms look for this kind of intense personal 212 identification, they appeal to salespersons as ‘worker-consumers’ who bring their extra-occupational enthusiasm for and knowledge of particular goods and brands to the job (Nickson et al. 2017, 698). The phenomenon of the worker-consumer has been described for upscale retail stores, where employers exploit personal identification to attract middle-class workers to jobs with poor employment conditions (Williams and Connell 2010), but we also find it in the low- to mid-scale shops of our sample. The following quote of a department store HR manager points to another use of the worker-consumer: If we take the fashion department, if a fashion-conscious lady or gentleman […] you feel, she has a flair, they know what they can put together for you, what type you are. You can have that without formal training in fashion. Beauty is the same: if someone never leaves the house without makeup or invests a lot of money in cosmetics, then she can sell that just as well as someone with training. The ‘flair’ of the worker-consumer fuels her passion for the job and at the same time compensates for the lack of vocational training. The privatization of skills—the fact that she acquired her competence informally in her private life—in turn makes her a cheaper worker. In the above quote, it remains implicit that the fashion and beauty conscious salesperson uses her flair not only to persuade customers but also employs it for the aesthetic labour of embodying the products and brands she sells (Nickson et al. 2017; Pettinger 2006). The quality of embodying the firm is deemed important in waitressing too (Monchâtre 2018), where the waiter or waitress should match the tar- get customers of a restaurant with respect to age or style. The interviewed managers in the catering industry also regard some vital skills like a flair for selling or friendliness as innate traits which cannot be trained. ‘You can learn everything else’, the HR manager of a large catering firm declares, ‘but you cannot learn friendliness’. In interactive service jobs so-called soft skills like friendliness and situa- tionally appropriate demeanour at the same time constitute technical skills in that they are indispensable for the core task of the respective jobs (Pettinger 2006). Nevertheless, as the examples of sales flair, friendliness or product knowledge show employers tend to discount such vital competencies as personal traits and incidental knowledge. The same contradiction between demanding certain qualities and denying their occupational nature applies to more obviously technical skills like numeracy or the ability to handle ma- chines and digital technology. Digital literacy, for example, is simply taken for granted. In the key competencies-discourse digital skills are regarded as crucial for employability and the digital illiteracy of the low-educated is pre- 213 sented as a pressing educational problem (Aepli et al. 2017; OECD 2015). In contrast, the firms of our sample stress that ‘we have employees who don’t speak German well and who don’t have a diploma, but we don’t have stupid workers’ (HR manager of a cleaning company). He and the vast majority of the interviewed firms argue that the digital technologies used in low-skilled jobs (tablets, touchscreens, handheld product scanners) are not more sophis- ticated than the smartphones and other electronic devices most people use in private life anyway.7 With this argument, a category of required skills is again reinterpreted as a privately acquired matter-of-course competence. On the other hand, managers see digital skills as a generational issue and believe eventual problems will fade away naturally with younger digital natives enter- ing the labour market. Until then, eventual deficits are tackled by simplifying the technology (see 3.) or by relying on the self-organization of teams, as the manager of a consumer goods factory explains: ‘If there are five people at a machine and one is maybe sixty and has respect for [using a tablet] then the 22-year-old does it and she does it in a completely natural way’. In this case, one worker’s uneasiness with digital technology is unproblematic as long as another team member can step in. From the manager’s point of view, the lack of skills of the individual worker is irrelevant, if the team as a whole can guarantee the smooth operation of production. Thus, the essentialization and privatization of skills create the interchangeability of workers which in turn disqualifies them as lacking distinguishing occupational qualities.8 The interchangeability of workers is in tension with the requirement of loyalty. Loyalty to the company is an important quality which firms try to gauge in the recruitment process, as the manager of a retail chain explains: ‘If you check the CV you want someone who has been long-standing somewhere so we will also have them for many years.’ Too many job changes are a signal for low quality: either for poor performance or for a tendency to disloyalty. The loyalty of workers guarantees a stable core workforce, which allows the use of the workers’ practical knowledge honed in years of experience on the job as well as profiting from well-functioning teams (Abel et al. 2014; 7 Digitalization does, however, exacerbate the issue of literacy and proficiency in the local language, because the technology necessitates some reading skills to understand task menus and instructions. The requirements for both, digital and language skills, are very low for entry-level jobs, but somewhat higher for workers with special tasks or with supervising functions. 8 Interchangeability is mentioned by some firms in a positive way as ‘polyvalence’: the ability to perform diverse tasks. Yet, sociologists argue that polyvalence adds to disqualification because workers cannot develop special skills and because it is not recognized as a skill in its own right (Cianferoni 2019; Demazière and Marchal 2018; Tranchant 2018). 214 H assler et al. 2019; Krenn et al. 2014). Except for the catering industry where high staff turnover is prevalent, the firms of our sample claim that a high percentage of their workers ‘stay forever’, i.e. for many years or decades. On the other hand, firms are interested in flexible employment to adapt to fluctuations in production. Hence, temporary work is widespread, especially in the construction and manufacturing firms of our sample where workers are hired almost exclusively through temporary staffing agencies. In addition to the advantages of flexibilization, temporary employment also provides a convenient practical test of the quality of the worker (Promberger 2012; Tranchant 2018). ‘If it doesn’t match, the temporary agency has to supply more people,’ remarks the manager of a food processing factory. In this way, workers can be returned on short notice like defective parts of a machine. The interchangeability of workers is thus a necessary condition for this kind of flexible use of a temporary workforce. Yet, there is also an element of sta- bility and loyalty within the flexible employment of interchangeable workers. First, ‘good’ temporary workers may be transferred to regular employment.9 Second, the firms often take on the same temporary workers during peak production periods for years. Because these workers have proved themselves and are already familiar with the tasks and work organization, there are no costs for initial training. In this way, the respective workers stand out from the indistinguishable pool of low-skilled labour and become valuable in a specific context. 9.5 Conclusions Low-skilled jobs are on the decline without vanishing completely and work- ers without formal qualifications have become a small minority within the labour force. They bear considerable risks and penalties like unemployment, low pay, and the social stigma of being ‘unskilled’ and unfit for more than ‘simple’ jobs. Yet, to a certain degree they are still in demand and—as we observe for our interview sample of low-skilled worker—some of them have surprisingly stable employment trajectories or even experience modest occu- pational advancement. This chapter examined the relation between low-skilled jobs and workers from the EC/SC-perspective of valorization and asked on what grounds firms value workers without occupational qualifications. Our 9 Only one manufacturing company of our sample does not offer permanent employment to low-skilled workers at all. The other firms use varying combinations of temporary workers alongside permanent employment contracts. Thus, we find a segmentation within the already peripheral labour market segment of low-skilled jobs (Köhler et al. 2017). 215 findings show that low-skilled workers can be employable despite (or because of ) the lack of formal qualifications and some of the key competencies deemed essential in discourses on the transformation of the labour market. In the eyes of employers, the important issue is the fit of the whole workforce with the firm’s specific coordination of production which, in turn, is subject to change according to the availability of suitable labour. The match between worker and work organization is both technical in terms of skills and productivity and social in terms of adapting to the firm as a community. Workers are judged for what they are able to do and for who they are. On the whole, their value rests on a compromise between the interchangeability of a nondescript category of workers lacking formal qual- ification or individual distinction and the stability of a loyal workforce with particular qualities that match the work organization and company culture of a specific firm. The good low-skilled worker fits the coordination of pro- duction like the cog in the wheel by supplying mostly physical capacities wherever it is not profitable or feasible to replace humans by machines, or where skilled workers are hard to find. In other words, workers are valuable within the rationalities of the market and the industrial convention. Accord- ing to market logic, the disadvantage of the worker (the low wage justified by the lack of qualification) is attractive for firms deliberately looking for low-skilled workers as cheap labour. Routine, mostly manual, and cognitively undemanding tasks are split off from skilled jobs and formatted as distinct ‘simple’ jobs for ‘everybody’. Workers in these jobs are then also valued ac- cording to the logic of the industrial convention for their ‘capacity to ensure normal operations’ (Boltanski and Thévenot 2006, 204) by performing the basic tasks that keep production running. Insofar the distinguishing feature of the individual low-skilled worker is fitting into the social order of the firm by proving motivation, identification, loyalty and the right blend of personal initiative and following orders, value is also constituted according to the domestic convention. However, we find a segmentation of the low- skilled workforce: a growing number of temporary workers is not admitted to stable membership in the company community, thus not only excluded from statutory rights of stable employment (Tranchant 2018) but also from the implicit moral obligations of employers for their staff beyond the work contract. Nevertheless, loyalty is still valued outside permanent membership: the fact that firms regularly re-employ the same temporary workers’ points to a compromise between market, industrial and domestic convention. Recurrent temporary employment ensures a pool of loyal, tried, and tested workers who are available and fully functioning ‘just on time’ whenever they are needed without causing costs while they are not. 216 The political and academic discourse on employability in face of the rap- idly changing world of work is very much future-oriented: the emphasis is on estimating trends in labour and skills demand and on policies to ensure the supply of employable labour. At the same time, political and academic discussions on employability tend to put a ‘singular focus on the individual and what might be termed their “employability skills”’ (Mc Quaid and Colin 2005, 205). In contrast, the firms of our sample are primarily concerned with the present and evaluate workers with regard to their collective workforce, within the context of the current requirements of the work organization. As we have seen, individual deficits, e.g. regarding language or digital skills may be compensated to a certain degree by the combined capacities of the team.10 Moreover, for some jobs firms cannot just replace less skilled individuals with higher-skilled ones because workers with formal qualifications would not accept these low-paid ‘simple’ jobs. Rather, employers use makeshift solutions to work around pervasive problems like, e.g. the low German language proficiency of a largely migrant workforce. They count on the collective capacities of teams—sometimes deliberately composing teams to that end, e.g. linguistically homogenous teams. And they integrate cognitive requirements into technology so tools and machines can be handled with a minimum of technical, language, or digital skills. In this way, the format of the coordination of production is altered to make jobs ‘simple’ and to adapt to the qualities of the available workforce. The other way around, in indus- tries with a pronounced skills shortage low-skilled workers are sometimes promoted to positions which are usually occupied by qualified workers. Both observations demonstrate that there are neither clear demarcations between ‘unskilled’ and skilled jobs nor between ‘unskilled’ and skilled workers. While the formats of production may be transformed in response to la- bour supply by creating simple jobs, firms are careful not to format valuable competencies as qualifications. As shown, some skills are not acknowledged as occupational qualification even though they are deemed essential for the job. Rather, they are naturalized as innate personal traits or privatized as taken-for-granted basic knowledge that ‘everybody’ can transfer from private life to the workplace. Although they have a value for firms and job applicants are screened for these skills, such seemingly extra-occupational qualities are not valorized in monetary terms or recognized as professional qualification. Competencies that are immediately useful for production may be trained on the job or with short courses, yet this kind of skills enhancement is rarely formatted by means of certificates, job titles, and occupational classifications. 10 Compensation is not tolerated for limited individual physical fitness, however (Nadai et al. 2021). 217 It can help the occupational advancement of individual workers within a particular firm, but it does not alter the distinction between qualified and unqualified workers and jobs. Workers without formal vocational training may be employable but they remain disqualified as ‘unskilled’. 9.6 References Abel, Jörg, Hartmut Hirsch-Kreinsen, and Peter Ittermann. 2014. Einfacharbeit in der Industrie. Strukturen, Verbreitung und Perspektiven. Berlin: Edition Sigma. Acker, Joan. 1990. Hierarchies, Jobs, Bodies. A Theory of Gendered Organizations. Gender and Society, 4(2): 139‒158. Aepli, Manuel, Vanessa Angst, Rolf Iten, Hansruedi Kaiser, Isabelle Lüthi, and Jürg Schweri. 2017. Die Entwicklung der Kompetenzanforderungen auf dem Arbeitsmarkt im Zuge der Digitalisierung. Bern: SECO. Atzmüller, Roland, Stefanie Hürtgen, and Manfred Krenn. 2015. Die zeitgemäße Arbeitskraft. Qualifiziert, aktiviert, polarisiert. Weinheim/Basel: Beltz Juventa. Benelli, Natalie. 2011. Nettoyeuse : comment tenir le coup dans un sale boulot. Zürich: Seismo. Bessy, Christian and Pierre-Marie Chauvin. 2013. The Power of Market Inter- mediaries: From Information to Valuation Processes. Valuation Studies, 1(1): 83‒117. Bessy, Christian, François Eymard-Duvernay, Guillemette de Larquier, and Ema- nuelle Marchal. 1999. Les institutions du recrutement. Approche comparative France/Grande Bretagne. Centre d’Etudes de l’Emploi, La Lettre, 60: 1‒10. https:// hal.archives-ouvertes.fr/hal-02137462/document (accessed June 15, 2020). BFS (Bundesamt für Statistik). 2020. Höchste abgeschlossene Ausbildung in der Schweiz. 30 January, 2020, https://www.bfs.admin.ch/bfs/de/home/ statistiken/bildung-wissenschaft/bildungsstand.assetdetail.11627129.html (accessed July 2, 2020). Bolli, Thomas, Christoph Breier, Ursula Renold, and Michael Siegenthaler. 2015. Für wen erhöhte sich das Risiko in der Schweiz, arbeitslos zu werden? KOF-Studien, 65: 1‒82. Boltanski, Luc and Laurent Thévenot. 2006. On Justification: Economies of Worth. Princeton: Princeton University Press. Boltanski, Luc and Laurent Thévenot. 2011. Die Soziologie der kritischen Kom- petenzen. In Rainer Diaz-Bone (Ed.), Soziologie der Konventionen. Grundlagen einer pragmatischen Anthropologie. 43‒68. Frankfurt am Main: Campus. Bosch, Gerhard and Claudia Weinkopf. 2011. «Einfacharbeit» im Dienstleis- tungssektor. Arbeit, 20(3): 173‒187. 218 Bourguignon, Annick and Eve Chiapello. 2005. The Role of Criticism in the Dynamics of Performance Evaluation Systems. Critical Perspectives on Accoun- ting, 16(6): 665‒700. Cianferoni, Nicola. 2019. Travailler dans la grande distribution: La journée de travail va-t-elle redevenir une question sociale? Zürich und Genf: Seismo. Crettaz, Eric. 2018. Working Poor in der Schweiz: Ausmass und Mechanismen. Social Change in Switzerland, 15, September 2018, https://www.socialchange switzerland.ch/wp-content/uploads/2018/09/Crettaz_Working_Poor_Ausmass_ Mechanismen_Sept2018.pdf (accessed September 12, 2018). Demazière, Didier and Emanuelle Marchal. 2018. La fabrication du travail non qualifié. Travail et emploi, 155‒156(3): 5‒30. Diaz-Bone, Rainer. 2018. Die «Economie des conventions»: Grundlagen und Ent- wicklungen der neuen französischen Wirtschaftssoziologie. 2nd ed. Wiesbaden: Springer Fachmedien Wiesbaden. Diaz-Bone, Rainer. 2009. Economie des Conventions. In Jens Beckert and Christoph Deutschmann (Eds.), Wirtschaftssoziologie, Kölner Zeitschrift für Soziologie und Spzialpyschologie, Sonderheft 49, 176‒193. Eymard-Duvernay, François. 2012. Du chômage keynésien au chômage d’exclu- sion. In François Eymard-Duvernay (Ed.), Epreuves d’évaluation et chômage. Toulouse: Octarès Editions, 9‒46. Eymard-Duvernay, François and Emmanuelle Marchal. 1997. Façons de recruter : le jugement des compétences sur le marché du travail. Paris: Éditions Métailié. Gesthuizen, Maurice, Heike Solga, and Ralf Künster. 2011. Context Matters: Economic Marginalization of Low-Educated Workers in Cross-National Perspective. European Sociological Review, 27(2): 264‒280. Hassler, Benedikt, Lea Widmer, Thomas Geisen, Nathalie Amstutz, Nico Scheide- gger, and Nadine Wenger. 2019. Arbeitsplätze ohne formale Qualifikations- anforderungen in der Schweiz. Rekrutierungsstrategien von Unternehmen und deren sozialpolitische Bedeutung. Zeitschrift für Sozialreform, 65(2): 147‒174. Hieming, Bettina, Karen Jaehrling, Thorsten Klaina, Achim Vanselow, and Clau- dia Weinkopf. 2005. Stellenbesetzung im Bereich «einfacher» Dienstleistungen. Abschlussbericht einer Studie im Auftrag des Bundesministeriums für Wirtschaft und Arbeit, BMWA-Dokumentation Nr. 550. Berlin: BMWA. Hieming, Bettina, Karen Jaehrling, Thorsten Klaina, Achim Vanselow, and Clau- dia Weinkopf. 2006. Der Arbeitsmarkt für «einfache» Dienstleistungstätig- keiten – Beschäftigungschancen für gering Qualifizierte? Sozialer Fortschritt, 55(8): 208‒214. Hilf, Ellen, Heike Jacobsen, Bärbel Meschkutat, and Katja Pohlheim. 2018. Be- rufsfachlichkeit im Einzelhandel – eine umkämpfte Ressource. Arbeits- und Industriesoziologische Studien, 11(1): 60‒75. 219 Holtgrewe, Ursula. 2015. Einfachdienstleistungen in der Fläche – die Unterseite der Tertiarisierung: Dienstleistungsgestaltung und -politik in schwierigem Gelände. Arbeits- und Industriesoziologische Studien, 8(1): 5‒20. Hürtgen, Stefanie. 2015. Qualifizierung und Polarisierung im transnationalen Raum der Lohnarbeit. Die zeitgemäß (un-)qualifizierte Arbeitskraft als betrieblich und gesellschaftlich konstruiertes Herrschaftsverhältnis. In Roland Atzmüller, Stefanie Hürtgen, and Manfred Krenn (Eds.), Die zeitgemäße Arbeitskraft. Qualifiziert, aktiviert, polarisiert. Weinheim/Basel: Beltz Juventa, 27‒124. Imdorf, Christian. 2010. ‘Emotions in the Hiring Process: How ‘Gut Feelings’ Rationalize Personnel Selection Decisions.’ In Barbara Sieben and Asa Wet- tergren (Eds.), Emotionalising Organizations and Organizing Emotions. New York: Palgrave Macmillan, 84‒105. Kleemann, Frank, Jule Westerheide, and Ingo Matuschek. 2019. Arbeit und Sub- jekt. Aktuelle Debatten der Arbeitssoziologie. Wiesbaden: Springer Fachmedien. Köhler, Christian, Stefan Schröder, and Simon Weingärtner. 2017. Arbeitsmärkte. Wirtschaftssoziologische Perspektiven. In Andrea Maurer (Ed.), Handbuch der Wirtschaftssoziologie. Wiesbaden: Springerl, 275‒305. Kozica, Arjan and Julia Brandl. 2015. Handling Paradoxical Tensions through Conventions: The Case of Performance Appraisal. German Journal of Human Resource Management, 29(1): 49‒68. Krenn, Manfred. 2015. An- und Ungelernte als Prototypen unzeitgemäßer Ar- beitskraft in der «Wissensgesellschaft». Stigmatisierung, Marginalisierung, Aktivierung. In Roland Atzmüller, Stefanie Hürtgen, and Manfred Krenn (Eds.), Die zeitgemäße Arbeitskraft. Qualifizert, aktiviert, polarisiert. Weinheim/ Basel: Beltz Juventa, 125‒194. Krenn, Manfred, Ulrike Papouschek, and August Gächter. 2014. Die Verbesse- rung der Erwerbschancen gering Qualifizierter durch lernförderliche Gestaltung einfacher Arbeit. Wien: FORBA Forschungs- und Beratungsstelle Arbeitswelt. Larquier, Guillemette de. 2016. Une approche conventionnaliste du marché du travail fondée sur le recrutement des entreprises. Paris: Université de Paris Ouest- Nanterre La Défense. https://hal.archives-ouvertes.fr/tel-01423201 (accessed January 12, 2017). Marchal. Emanuelle. 2015. Les embarrass des recruteurs. Enquête sur le marché du travail. Paris: Editions EHESS. Marchal, Emanuelle and Géraldine Rieucau. 2010. Le recrutement. Paris: La Découverte. McQuaid, Ronald and Lindsay Colin. 2005. The Concept of Employability. Urban Studies, 42(2): 197‒219. Monchatre, Sylvie. 2018. Embacuher ou habiliter ? Recrutement et espaces de (non-)qualification dans l’hôtellerie-restauration. Travail et emploi, 155–156(3): 93–114. 220 Murphy, Emily and Daniel Oesch. 2018. Is Employment Polarisation Inevitable? Occupational Change in Ireland and Switzerland, 1970‒2010. Work, Employ- ment and Society, 32(6): 1‒19. Nadai, Eva, Alan Canonica, Anna Gonon, Fabienne Rotzetter, and Martin Lengwi- ler. 2019. Werten und Verwerten. Konventionen der Beschäftigung von Menschen mit Behinderungen in Wirtschaft und Wohlfahrtstaat. Wiesbaden: Springer VS. Nadai, Eva, Anna Gonon, Robin Hübscher, and Anna John. 2021. The Social Organization of Work Incapacity. Incapacities in the Swiss Social Insurance System and in the Workplace. Historical Social Research, 46(1): 160-180 Nathani, Carsten, Pino Hellmüller, Corina Rieser, Oliver Hoff, and Sujetha Nesa- rajah. 2017. Ursachen und Auswirkungen des Strukturwandels im Schweizer Arbeitsmarkt. Bern: SECO. Arbeitsmarktpolitik, 46. November, 2017, https:// www.seco.admin.ch/seco/en/home/Publikationen_Dienstleistungen/ Publikationen_und_Formulare/Arbeit/Arbeitsmarkt/Informationen_Arbeits marktforschung/ursachen_auswirkungen_strukturwandels_arbeitsmarkt.html (accessed July, 20, 2020). Nickson, Dennis, Robin Price, Hazel Baxter-Reid, and Scott A. Hurrell. 2017. Skill Requirements in Retail Work: The Case of High-End Fashion Retailing. Work, Employment and Society, 31(4): 692‒708. OECD. 2015. Adults, Computers and Problem Solving: What’s the Problem? Paris: OECD Publishing. OECD. 2016. Getting Skills Right. Assessing and Anticipating Changing Skill Needs. Paris: OECD Publishing. OECD. 2019. The Survey of Adult Skills: Reader’s Companion, Third Edition. Paris: OECD Publishing. Pettinger, Lynne. 2004. Brand Culture and Branded Workers: Service Work and Aesthetic Labour in Fashion Retail. Consumption Markets & Culture, 7(2): 165‒184. Pettinger, Lynne. 2006. On the Materiality of Service Work. The Sociological Review, 54(1): 48‒65. Promberger, Markus. 2012. Topographie der Leiharbeit: Flexibilität und Prekarität einer atypischen Beschäftigungsform. Berlin: Edition Sigma. Promberger, Markus, Ulrich Wenzel, Sabine Pfeiffer, Anne Hacket, and Andreas Hirseland. 2008. Beschäftigungsfähigkeit, Arbeitsvermögen und Arbeitslo- sigkeit. WSI-Mitteilungen, 61(2): 70‒76. Salais, Robert. 2001. A Presentation of the French “Economie des conventions”. Application to Labour Issues. Work seminar of the ESCR-Center of Business Research at the University of Cambridge. United Kingdom: University of Cam- bridge. March, 2001, https://halshs.archives-ouvertes.fr/halshs-00430648 (accessed March 20, 2017). 221 Sardadvar, Karin, Angelika Kümmerling, and Darina Peycheva. 2015. ‘Gender-Typ- ing of Low-Wage Work: A Comparative Analysis of Commercial Cleaning in Austria, Domiciliary Elderly Care in Germany and the Waste Sector in Bulgaria.’ In Ursula Holtgrewe, Vassil Kirov, and Monique Ramioul (Eds.), Hard Work in New Jobs. The Quality of Work and Life in European Growth Sectors. London: Palgrave Macmillan UK, 208‒230. Smith, Vicki and Esther B. Neuwirth. 2009. Temporary Help Agencies and the Making of a New Employment Practice. Academy of Management Perspectives, 23(1): 56‒72. Solga, Heike. 2008. ‘Lack of Training—The Employment Opportunities of Low- Skilled Persons from a Sociological and Micro-Economic Perspective.’ In Karl Ulrich Mayer and Heike Solga (Eds.), Skill Formation—Interdisciplinary and Cross-National Perspectives. New York: Cambridge University Press, 173‒204. Thévenot, Laurent. 1984. Rules and Implements: Investment in Forms. Social Science Information, 23(1): 1‒45. Thévenot, Laurent. 2001. Organized Complexity. Conventions of Coordination and the Composition of Economic Arrangements. European Journal of Social Theory, 4(4): 405‒425. Tranchant, Lucas. 2018. L’intérim de masse comme vecteur de disqualification professionnelle. Le cas des emplois ouvriers de la logistique. Travail et emploi, 155–156(3) : 115–140. Vallas, Steven Peter. 1990. The Concept of Skill: A Critical Review. Work and Occupations, 17(4): 379‒398. Waldinger, Roger and Michael I. Lichter. 2003. How the Other Half Works: Immigration and the Social Organization of Labor. Berkeley: University of California Press. Wickham, James. 2011. Low Skill Manufacturing Work: From Skill Biased Change to Technological Regression. Arbeit, 20(3): 224‒238. Williams, Christine L. and Catherine Connell. 2010. “Looking Good and Sound- ing Right”: Aesthetic Labor and Social Inequality in the Retail Industry. Work and Occupations, 37(3): 349‒377. Wotschack, Philip. 2017. Unter welchen Bedingungen bilden Betriebe an- und ungelernte Beschäftigte weiter? Zeitschrift für Soziologie, 46(5): 362‒380. Wotschack, Philip and Heike Solga. 2014. Betriebliche Weiterbildung für be- nachteiligte Gruppen. Förderliche Bedingungskonstellationen aus instituti- onentheoretischer Sicht. Berliner Journal für Soziologie, 24: 367‒395. 222 10 Éloignement de l’emploi : trois dimensions du registre de la honte Jean-Pierre Tabin, Monika Piecek, Isabelle Probst et Céline Perrin1 La honte. Cela ne crie pas. C’est un refroidissement. (Difficultés, Henri Michaux 1930) He had joined the brigade of the damned, he told himself, and from now on he would be looked upon as one of those crippled, distorted people who no longer counted as full-fledged members of the human race. (4321, Paul Auster 2017) 10.1 Introduction Les allocataires de prestations de l’État social rendent fréquemment visibles dans l’espace social des sentiments tels que la honte, la gêne, la culpabilité ou l’embarras (De Gaulejac 1996 ; Tabin et al. 2010). Cette expression d’affects – par un discours, une hexis corporelle – a par exemple lieu lors de contact avec des agent·e·s de l’État social (Driessens 2010). Nous l’avons nous-même régulièrement entendue dans les entretiens de recherche que nous avons menés avec des personnes en contact avec l’assurance invalidité suisse (AI), un dispositif social public habilité à décider si une personne est « invalide », « partiellement invalide » ou « valide ». Précisons que l’invalidité est définie comme une « diminution [permanente ou à long terme] de l’ensemble ou d’une partie des possibilités de gain […] [qui] résulte d’une atteinte à sa santé physique, mentale ou psychique et [qui] persiste après les traitements et les mesures de réadaptation exigibles » (Loi fédérale sur la partie générale du droit des assurances sociales, art. 7 et 8). Si le retour à une « capacité de gain » 1 Une version antérieure de ce chapitre a paru en anglais : Tabin, Jean-Pierre, Piecek, Monika, Perrin, Céline, & Probst, Isabelle. 2019. Three Dimensions in the Register of Shame. Review of Disability Studies. An International Journal 15(3), https://rdsjournal. org/index.php/journal/article/view/871 (4.3.2021). 223 semble possible, l’AI peut proposer aux assuré·e·s des programmes de « réa- daptation », organisés par des prestataires (Piecek et al. 2017) qui poursuivent des objectifs de retour sur le marché de l’emploi, donc de re-marchandisation (Holden 2003 ; Offe 1984). Des pensions (partielles ou complètes, selon le degré d’invalidité) sont versées aux personnes catégorisées invalides. Du point de vue sociologique, cette pratique du registre de la honte par des personnes dont la capacité de gain est diminuée en raison d’atteintes à la santé et dont l’avenir dans l’emploi est devenu hypothétique n’est guère sur- prenante, et ce pour deux raisons intriquées. D’une part, parce que la Suisse comme les autres sociétés contemporaines est traversée par des logiques qui mettent en valeur les capacités ; elles dévalorisent les personnes qui ne réus- sissent pas, pas complètement ou plus à correspondre aux normes dominantes (Stiker 2005), notamment en matière d’emploi durant l’âge dit « actif ». La discrimination sociale basée sur les capacités – le capacitisme – les place en situation inférieure (Campbell 2009 ; Goodley 2014). Les personnes qui ne répondent pas aux attentes concernant la performance sont en effet systémati- quement décrites par la négative, comme incapables, handicapées, infirmes ou invalides. Il n’est guère étonnant que cette expérience les affecte. La référence à la honte, la gêne, la culpabilité ou l’embarras est une réaction logique en situation de rupture avec la norme de l’emploi, d’autant plus en interaction avec des personnes qui se considèrent et sont considérées comme « capables ». D’autre part, les nombreux changements récents des politiques suisses en matière d’invalidité, avec leur litanie de discours sur l’abus de prestations sociales (Ferreira et Frauenfelder 2007 ; Frauenfelder et Togni 2007), ont mis l’accent sur la responsabilité des personnes qui font valoir leurs droits sociaux. Ce sont des injonctions à se rétablir, à améliorer son employabilité ou à s’ouvrir à une plus grande palette de professions, ou encore à cesser de compter sur les prestations de l’assurance invalidité (Piecek et al. 2019 ; Probst et al. 2016). Ce type d’encadrement moral, aujourd’hui dominant dans le discours politique du Nord global, a joué un rôle central dans la mise en œuvre de programmes d’activation visant à réintégrer rapidement les personnes invalides, au chômage ou à l’assistance au marché du travail (Parker Harris et al. 2014 ; Yates et Roulstone 2013). Ces programmes visent à modifier les comportements individuels et sont définis comme des moyens de surmonter la « culture de l’assistance » (OECD 2009). Les personnes qui demandent des prestations ne sont dès lors pas considérées comme des victimes de l’insécurité socio-économique liée à leur position sociale, mais comme des risques financiers pour les régimes sociaux. Comme le fait remarquer Skeggs (2004, 82), ‘[under liberal policy frameworks], those who cannot perform their state-defined “duty” are thus morally suspect’. Les personnes 224 que nous avons rencontrées nous ont donc logiquement parlé des émotions négatives qu’elles ont éprouvées, et parfois disent toujours ressentir en raison de leur éloignement de l’emploi et de leur confrontation aux dispositifs de l’assurance invalidité. À l’analyse, on s’aperçoit toutefois que la pratique du registre de la honte dans le discours des personnes rencontrées en entretien est plus complexe qu’il n’y paraît. Dans ce chapitre, nous allons montrer que si le recours au registre de la honte ne remet pas frontalement en question les hiérarchies capacitistes, il déplace les significations attachées à l’invalidité. Nous allons le voir à l’aide d’une analyse faisant ressortir trois dimensions du registre de la honte (le positionnement social, le jugement social et l’effet-miroir). Mais d’abord, nous allons faire un rapide état de la littérature sociologique sur la honte et expliquer notre méthodologie. 10.2 Aspects théoriques Au cours de la dernière décennie, la littérature scientifique traitant de la honte s’est considérablement enrichie dans le contexte de ce que certain·e·s auteur·e·s appellent le « tournant affectif » des sciences sociales (Clough et Halley 2007 ; Lordon 2013). Les différentes analyses se situent souvent au carrefour de théories psychologiques et sociologiques. Remarquons d’abord que la honte est définie de manière plus ou moins spécifique. Si certain·e·s auteur·e·s tiennent à la distinguer notamment de la culpabilité, de l’humiliation ou de l’embarras (Biddle 1997 ; Probyn 2005), d’autres soulignent au contraire la nature perméable des expériences émo- tionnelles et mettent en garde contre le risque de réification des catégories lié à des définitions trop circonscrites (Barrett 2016). Selon ces scientifiques, les émotions coexistent et s’interpénètrent de manière dynamique en fonction du contexte de l’interaction (Kosofsky Sedgwick 1993 ; Munt 2008). Pour Scheff (2000, 96), la honte est dès lors ‘a large family of emotions that includes many cognates and variants, most notably embarrassment, humiliation, and related feelings such as shyness, that involve reactions to rejection or feelings of failure or inadequacy’. Nous adoptons dans ce chapitre ce point de vue qui permet de rendre compte de l’expérience des personnes que nous avons interrogées et d’inclure tout l’éventail des termes qu’elles ont utilisés. D’un point de vue sociologique, la honte est un affect lié à la position sociale des individus et aux expériences qui en découlent (Loveday 2016 ; Skeggs 1997). Cet affect est façonné par l’histoire et la culture, mais également par les normes et valeurs sociales dominantes. À cet égard, comme le relève Ahmed 225 (2004, 107), la honte peut être appréhendée comme ‘the affective cost of not following the scripts of normative existence’. L’expérience des personnes que nous avons rencontrées et qui sont en contact avec l’AI correspond à cette définition, car l’assurance n’intervient pour les adultes que si une atteinte à la santé physique, psychique ou mentale est constatée qui risque d’obliger une personne à quitter définitivement le monde de l’emploi, ou à de ne plus pouvoir effectuer les tâches domestiques. Il y a donc une déviation par rapport au cours considéré comme normal de l’existence, le retrait de l’emploi pour d’autres raisons que « faire famille » n’étant normalisé qu’à partir de l’âge de la retraite et l’incapacité des adultes n’étant socialement attendue que dans le grand âge (en 2016, le taux d’activité net de la population dite « active » de quinze à soixante-quatre ans est de 81,3 % en Suisse2). Nous allons analyser dans la suite de ce chapitre comment cette déviation est expérimentée. D’autres perspectives nous permettent d’approfondir notre compréhension de la honte. La perspective interactionniste a depuis ses textes fondateurs (Goffman 1973 ; Mead 1963) mis l’accent sur le caractère réflexif et inter- personnel de cet affect. Elle met en lumière que la honte est liée à la fois à un jugement sur soi et au regard d’autrui, ou à l’anticipation de ce regard, ce qui signifie qu’elle implique non seulement une conscience des normes sociales, mais encore du jugement des autres. Elle résulte dès lors d’une appréciation morale dans laquelle la personne imagine le regard disqualifiant que d’autres (réel·le·s ou imaginé·e·s, présent·e·s ou absent·e·s) portent sur elle. Cela signifie que les actrices et acteurs sociaux répondent aux situations autant en fonction d’interactions sociales que de leurs anticipations ou projections. Les personnes qui sont en contact avec l’AI, dans un contexte politique marqué depuis bientôt trente ans par de multiples débats sur cette assurance3 largement relayés par les médias, savent que le regard d’autrui sur leur situation n’est pas uniquement marqué par la bienveillance, et qu’il peut être disqualifiant. Il faut rappeler ici que le mode de financement de l’AI, fondé sur un taux de cotisation fixe et sur l’absence de fortune propre (à l’exception d’un fonds de compensation dont le rôle est de garantir des liquidités) est précaire (Fracheboud 2015) et que « les personnes considérées comme invalides sont […] désignées [depuis le milieu des années 1990] comme responsables de l’augmentation des coûts » (Probst et al. 2015, 108). Dans une perspective interactionniste, ces débats ont des effets sociaux et peuvent produire de la honte, puisque le discours 2 Source : Office fédéral de la statistique, 03 – Travail, rémunération / Structure de la population active / Taux d’activité net / 2016, www.atlas.bfs.admin.ch/maps/13/ fr/12954_8756_127_119/21287.html (03.03.2021). 3 Six révisions majeures de la loi mises en chantier, dont quatre ont abouti a un chan- gement législatif. 226 social sur les personnes considérées comme invalides, comme celui sur les pauvres, tend à expliquer leur situation, et par extension celle de l’assurance, par des défaillances individuelles. Dans ce contexte, l’argument du mérite, ou en l’occurrence de son absence, se conjugue avec le discours sur le parasitisme : les récipiendaires sont en effet parfois définis par les médias, le public, les politiques ou les institutions sociales comme des « paresseux », des « abuseurs », des « profiteurs », comme à l’occasion de la votation populaire sur la 5e révision de l’AI (Ferreira 2015). Ces discours stigmatisants légitiment le contrôle, récemment renforcé avec l’introduction d’une révision systématique des pensions et avec l’obligation faite aux assuré·e·s de « prendre toutes les mesures pour éviter de bénéficier des prestations de l’assurance » (Despland 2012, 77). Comme s’en vante le gouvernement helvétique, depuis « le 1er janvier 2008, l’AI mène une lutte active […] contre les abus dans l’assurance […] qui peut se décomposer en quatre phases : l’identification des cas suspects, l’enquête et les examens approfondis, la surveillance en tant que dernier moyen d’enquête et de constitution de la preuve et, enfin, l’application du droit des assurances et de mesures pénales » (Conseil fédéral 2017, 4). Les autorités participent depuis de nombreuses années à diffuser l’idée qu’un nombre potentiellement significatif d’invalides perçoit ou tente d’obtenir des prestations indues, jetant un discrédit sur l’ensemble des récipiendaires de pensions. Cela explique en partie le contexte qui fait qu’un lien est réalisé entre honte et invalidité. La honte n’est alors pas qu’un affect, c’est un outil moral (Chase et Wal- ker 2013). À l’image des ‘politics of resentment’, elle forme ‘an exclusionary emotional and social framework that traps minority identities and people experiencing multiple deprivations in its belief system and practices’ (Hughes 2015, 996). En ce sens, la honte, l’humiliation comme le dégoût permettent de maintenir la frontière entre le « normal » et l’« anormal » en reconstituant ce qui est acceptable, comme le démontre Moore (2016) à l’exemple des processus d’humiliation des personnes ayant transgressé les normes de genre. Cette politique des émotions (Skeggs 1997), qui est également une économie morale (Skeggs 2009), confère de la valeur à certaines personnes aux dépens d’autres. Dès lors, la honte agit comme un régulateur social qui (re)produit des hiérarchies sociales et des inégalités. C’est un pouvoir subtil incitant à l’auto- régulation des comportements et à la normalisation des conduites (Baker 2013 ; Creed et al. 2014), parce que ‘we are not merely “done in” by those who have power ; rather, we are all “doing” power’ (Fox 2014, 343‒344). Si Scheff (2000) soutient que la honte est un ciment qui unit la société, Walker (2014, 191) estime qu’elle ‘might be better described as cement reinforcing 227 structures of inequality and perpetuating poverty’. Cela signifie que la honte naturalise les stratifications sociales : elle fait « vivre sur le mode du péché originel et de l’indignité essentielle des différences qui […] sont le produit de conditionnements sociaux » (Bourdieu et Delsaut 1975, 36), ce qui a pour conséquence que les personnes qui la vivent « contribuent, souvent à leur insu, parfois contre leur gré, à leur propre domination en acceptant tacitement les limites imposées » (Bourdieu 1998, 60). C’est le « sens de sa place » sociale qui s’éprouve : la honte de celle ou de celui qui se sent déplacé·e ou l’aisance associée au sentiment d’être à sa place (Bourdieu 1997). La honte reflète la conscience qu’une personne a de sa position sociale, c’est ‘an embodied sense of self-judgment’ (Barrett 2016, 5). En résumé, la honte est une dimension d’un habitus, liée à un positionne- ment dans l’espace social. Elle se développe dans les contextes qui considèrent comme inférieures certaines positions sociales, l’une des logiques inhérentes à ces processus étant la hiérarchisation des individus en fonction des capacités. Lorsqu’elle est incorporée par les personnes occupant ces positions sociales, elle devient un puissant outil de reproduction des hiérarchies. Toutefois, la recherche poststructuraliste a montré que comme la honte construit l’individu comme sujet (Fullagar 2003 ; Munt 2008), elle pourrait être une ressource pour questionner les rôles sociaux. La honte, selon ce point de vue, n’est pas seulement une émotion négative (Chase et Walker 2013), voire débilitante (Weiss 2010), elle affecte de manière plus ambivalente la représentation de soi. Nous allons voir ce qu’il en est dans la suite de ce chapitre. 10.3 Méthodologie Ce chapitre s’appuie sur trente-trois récits de personnes ayant participé à des programmes de réadaptation professionnelle de l’AI suisse récoltés entre février 2016 et janvier 2017 dans le cadre du projet Le nouveau paradigme de l’assurance invalidité suisse financé par le Fonds national suisse de la recherche scientifique. Le but de la recherche était de comprendre l’expérience que font les personnes concernées des récentes réformes de l’AI (Piecek et al. 2019). Nous avons récolté une large palette d’expériences de réadaptation avec l’AI de manière à respecter les principes posés pour la sélection d’échantillons qualitatifs de cas multiples (Glaser et Strauss 1967 ; Pires 1997). Au cours du processus de recrutement des participant·e·s, nous avons contacté près d’une centaine d’organisations comme des associations de personnes malades ou en situation de handicap, des fondations, des groupes de soutien, des syndicats, des structures résidentielles psychosociales ou des services sociaux, et leur avons 228 demandé, dans le respect des critères éthiques, de relayer nos sollicitations d’interviews. Malgré le nombre de contacts amorcés, nous avons éprouvé certaines difficultés à recruter des participant·e·s. Nous avons eu des refus à cause de situations personnelles particulièrement difficiles (en termes de santé ou de conflit avec l’AI), ou de la crainte que les informations transmises lors de l’entretien reviennent aux oreilles des agent·e·s de l’AI. Nous avons rencontré treize femmes et vingt hommes âgé·e·s de vingt à soixante-quatre ans (treize avaient moins de quarante ans, vingt entre quarante et soixante-quatre ans). Ces personnes avaient l’expérience de différents dia- gnostics d’atteintes à la santé (surdité, douleur chronique, dépression, cancer, épuisement professionnel). Elles étaient à diverses étapes de leurs contacts avec l’AI : certaines participaient encore à des programmes de réadaptation, d’autres les avaient terminés ; certaines étaient retournées à l’emploi, d’autres non. La durée de leurs contacts avec l’AI variait de quelques mois à plus de vingt ans. Le canevas d’entretien semi-structuré a été conçu pour mettre l’accent sur l’évaluation des interventions de l’AI afin d’éviter de reproduire la structure des questionnaires de l’AI sur les troubles de santé et l’expérience profession- nelle. Bien qu’aucune des questions n’ait porté explicitement sur les émotions, onze personnes ont spontanément utilisé le terme « honte » pour décrire leur expérience, et vingt-neuf ont fait référence à cet affect en utilisant des mots tels que « stigmatisant », « dur », « difficile », « pénible », « humiliant », etc. Ces mises en forme verbales de la honte comprenaient également des expressions métaphoriques, comme : « Ce n’est pas quelque chose qu’on crie sur les toits ! » Tous les entretiens ont été enregistrés, transcrits et rendus anonymes. Nous avons identifié trois dimensions du registre de la honte que nous reprenons dans la suite de ce chapitre (le positionnement social, le jugement social et l’effet-miroir). Au moins une des dimensions apparaît dans 29 des 33 entretiens réalisés. La majorité des personnes ont évoqué deux dimensions (17 personnes), huit  personnes en ont évoqué trois et quatre personnes n’en ont évoqué qu’une seule. Nous avons constaté que des attributs comme l’âge, le niveau de scolarité ou le type de diagnostic ne sont pas associés à un type particulier de discours sur la honte. Autant les personnes qui tirent un bilan positif de l’intervention de l’AI que celles qui considèrent que l’action de l’assurance a été inutile ou a péjoré leur état de santé relatent des expériences de honte. Le seul élément qui semble jouer un rôle est l’appartenance de sexe : nous avons en effet constaté que parmi les personnes ayant évoqué les trois dimensions du registre de la honte, les femmes sont majoritaires (6 personnes sur 8). 229 10.4 Première dimension de la honte : le positionnement social Les personnes que nous avons rencontrées nous ont expliqué avoir, après l’atteinte à la santé qui les a éloignées de l’emploi, redouté le contact avec l’AI à cause du (re)positionnement social qu’il implique. Dans le cours normal d’une vie, la grande majorité des personnes n’est en effet jamais en contact avec cette assurance sinon pour payer des cotisations, qui sont obligatoires. Devoir s’y confronter est une expérience qualifiée d’« impossible », d’« inconce- vable » ou de « choquante ». C’est ce que relève Martine (infirmière4, 47 ans) : « Au début, je ne voulais pas y aller parce que je me suis dit : ‹ Je n’ai rien à voir avec l’invalidité, je suis juste en burn-out, j’ai besoin d’un peu de temps pour m’en remettre ›. » Ivana (en reconversion professionnelle dans le secteur administratif, 34 ans) confirme que la confrontation à l’AI a été une remise en question de sa valeur sociale : « Je l’ai mal pris. Je me suis sentie inutile. […] J’avais l’impression d’être… pas vraiment une moins que rien, mais presque. » Décrit souvent à travers des métaphores spatiales descendantes (chuter, tomber, retomber, atterrir), ce repositionnement parfois soudain à une place inférieure de l’espace social est porteur d’une forte charge symbolique. Comme l’a souligné Probyn (2005 : x), la honte qui en découle ‘puts one’s self-esteem on the line and questions our value system’. L’entrée en contact avec l’AI est une rencontre avec un lieu de pouvoir (Bourdieu 2012). L’AI à travers ses agent·e·s a le monopole d’attribuer un statut (in/valide), de catégoriser (valide, partiellement ou totalement invalide) et de juger (peut ou non recevoir une pension, participer à des programmes de réadaptation, etc.), ce qui fait de cette institution la garante d’un ordre. Les personnes s’adressant à l’AI savent qu’elles sont confrontées à une institution publique qui possède la légitimité de les placer sur une échelle d’in/capacité, et que cette rencontre va influer durablement sur leur position sociale : elle va leur ouvrir – ou non – des droits à des services sociaux et financiers qui vont peser sur leur qualité de vie. Les individus sont doublement démunis dans cette confrontation où se joue leur futur dans l’emploi. D’une part, ils n’ont pas d’autre choix, vu leur positionnement nouveau en regard des normes de capacité, que de s’adresser à l’AI pour trouver de quoi vivre en dehors de l’emploi (la pension complète peut permettre d’atteindre le minimum vital) ou pour recouvrer, via des programmes de réadaptation, une capacité de gain par l’emploi. « Là, je suis obligée de le faire, […] c’est juste que je suis obligée », souligne Ivana (en 4 Les professions indiquées sont celles que les personnes occupaient au moment de l’entretien. 230 reconversion professionnelle dans le secteur administratif, 34 ans). Comme le relève Brigitte (stagiaire dans un secrétariat, 55 ans) : « Personne n’a envie d’être à l’AI […], c’est que ça fait partie d’un traitement, comme l’insuline quand on est diabétique. » Pour Sonia (rentière AI, 46 ans), « une personne sensée ne la ferait pas, cette demande ». D’autre part, l’AI leur signifie systématique- ment qu’elles ne sont qu’un individu face à une institution. Par exemple, les systèmes de classification de l’assurance ne leur sont pas communiqués. Les courriers de l’AI sont écrits dans un langage juridique abscons. Les agent·e·s changent et l’institution est difficile à joindre. Marie (en attente de décision de l’AI, 48 ans) décrit ainsi comment sa honte a été nourrie : Depuis 2012, il n’y a plus personne. Plus personne ne s’occupe de mon dossier. […] Et c’est terrible ça, vous avez l’impression que vous n’existez plus. Vous n’êtes plus bonne à rien. Outre cette confrontation concrète avec le pouvoir, se rendre à l’office AI signifie pour les personnes que nous avons rencontrées institutionnaliser un éloignement du standard de la normalité, donc potentiellement perdre définitivement le statut d’adulte exerçant une activité productive, capable, disposant d’un corps ou d’un psychisme lui permettant d’être à la hauteur des exigences du marché de l’emploi. Si être dans un rapport de dépendance à un employeur semble « normal », la dépendance financière à l’État a pour conséquence de produire de la honte et un sentiment d’exclusion (Fraser et Gordon 1994 ; Young 2003). Cette différence vient sans doute du fait que si l’AI peut dédouaner une personne de l’obligation d’emploi pendant l’âge dit « actif » en fournissant une pension, elle institue un statut vécu comme marqueur de moindre valeur sociale. Comme le dit Jérôme (en formation de conseiller de vente, 46 ans), c’est « un peu difficile de dire que voilà, qu’on est à l’AI, qu’on est comment on dit ça, un peu rabaissé si vous voulez ». Brigitte (stagiaire dans un secrétariat, 55 ans) relève que se « retrouver là, à l’AI, c’est horrible », tandis que Marie (en attente de décision de l’AI, 48 ans) explique : « Quand je me vois maintenant, personne ne pourrait imaginer que j’étais infirmière. […] Vous imaginez jusqu’où je suis tombée ? C’est terrible. » Cela signifie que le dédouanement de l’obligation d’emploi que permet la recon- naissance du statut d’invalide n’est pas vécu sur le mode de l’affranchissement des contraintes du monde du travail, mais comme un destin funeste. Jean, 48 ans, ancien aide-soignant et désormais pensionné AI, le relève clairement : Je l’ai très mal pris quand ils m’ont mis à 100 pour cent à l’AI. J’étais fâché. […] [L’agent de l’AI] m’a fait comprendre que c’était pour mon bien […] et […] qu’il n’y avait pas de honte. […] Il ne se rend 231 pas compte que ça peut peser sur mes enfants. […] les étiquettes, ça reste, […] c’est très lourd, les étiquettes. La honte surgit également à cause du processus complexe de dévoilement que le contact avec l’AI implique. Au cours de l’enquête de l’AI, les personnes doivent en effet faire reconnaître que les atteintes à la santé dont elles souffrent diminuent leur capacité de gain. Cela signifie que le contact avec l’assurance implique d’exposer ses problèmes de santé et leurs effets. Cette présentation de soi à travers ses limitations est source de honte. C’est par exemple l’expérience relatée par Helmut (en attente de décision de l’AI, 25 ans) : Et je lui [à son conseiller AI] ai vraiment expliqué tout, tout, tout, tout. Et même que pour être au bureau à 8 heures, je dois me lever à 5 heures […] Je lui ai expliqué comment faire pour aller aux toilettes quand on est à moitié paralysé, on ne peut pas aller aux toilettes en 5 minutes, ça prend quarante minutes tous les matins. C’était un peu gênant d’en parler […]. Donc, je lui ai fait comprendre […] que je devais faire un massage au ventre pour stimuler les intestins. Enfin, que des trucs comme ça et il était choqué le gars, il n’en revenait pas. Lui, il pensait que j’étais juste quelqu’un qui était assis et puis que mon corps fonctionnait comme quelqu’un d’autre. De nombreuses recherches sociohistoriques ont montré que le corps handi- capé a été socialement constitué comme monstrueux, excessif, contaminé, dangereux ou sans défense (Davis 1995 ; Mitchell et Snyder 2000 ; Shildrick 2002). Le processus de l’AI oblige les potentiel·le·s allocataires à s’identifier à ce « corps handicapé » et à exposer leurs atteintes à la santé afin d’être admissibles aux prestations. Le fait de décrire et de reconnaître les façons dont leur activité est limitée par leur état physique ou mental force les per- sonnes confrontées à des instruments d’évaluation à objectiver leur corps et à se (re)construire comme handicapées (Reeve 2012). L’AI contribue dès lors à une construction du corps comme « anormal », « déviant » et « honteux ». La capacité de travail d’une personne est évaluée en fonction de son écart avec le point de référence « normal » ou « idéal » des politiques sociales. Ces processus font partie intégrante de la conception et de la mise en œuvre de politiques sociales fondées sur l’identification, la catégorisation, la mesure et l’élimination de l’« anormal ». Cette situation n’est pas particulière à la Suisse, on la retrouve par exemple au Royaume-Uni où l’obligation de prouver que le corps est improductif et divergent de la norme a été renforcée au cours de ces quinze dernières années (Goodley et al. 2014). Ce déplacement et l’identification à une position dominée dans un sys- tème qui valorise les capacités a des conséquences multiples, car le passage à 232 l’AI peut également influer sur la capacité de jouer d’autres rôles sociaux, par exemple ceux liés à la parentalité ou au rôle traditionnel masculin. Olivier (maître socioprofessionnel, 46 ans) parle en ces termes de ce qu’il lui semblait rester de sa capacité à séduire : J’avais 28 ans quand j’ai commencé à toucher la rente, donc du coup un homme 28 ans, jeune, pas marié, je me disais : « Mais attends, quelle femme va tomber amoureuse de toi ? » Enfin je veux dire, mais ça, c’était moi. Je me suis réduit au départ avec des idées : « Mais attends, tu es cuit là. Tu es cuit, tu n’as plus de valeur. » Dans vingt-sept des entretiens que nous avons menés, nous trouvons cette première dimension du registre de la honte : l’assignation à une position sociale dévalorisée, donc, comme nous l’avons relevé dans la partie théorique de ce chapitre, une déviation par rapport au cours considéré comme normal de l’existence qui oriente la personne vers une place inférieure dans la hiérarchie capacitiste. Nos données montrent le rôle joué par l’État dans ce processus d’assignation à la position subalterne et donc dans la production de la honte. 10.5 Deuxième dimension de la honte : le jugement social La deuxième dimension du registre de la honte, présente dans vingt de nos entretiens, est liée au discours généralisé sur l’abus de prestations de l’AI que nous avons évoqué plus haut. Aline (auxiliaire de crèche, 32 ans) relate que lorsqu’elle a contacté l’AI, elle a « culpabilisé ». Sonia (rentière AI, 46 ans) explique « avoir ce sentiment : oui, je suis malhonnête, je profite du système. Oui, d’être un escroc », tout comme Ivana (en reconversion professionnelle dans le secteur administratif, 34 ans) qui a « ce sentiment que les gens te regardent un peu bizarrement. Tu as l’impression d’être comme un profiteur ». Martine (infirmière, 47 ans) fait part d’une expérience similaire : « Voilà, je suis arrivée en 2013 en Suisse, je me suis retrouvée en burn-out, du coup, je suis à l’AI. […] Mais qu’est-ce qu’on va penser de moi ? » Ces expériences découlent du discours moral sur l’abus de prestations de l’AI ainsi que de la valeur de l’emploi. La conduite visée par la rhétorique de l’abus est la malhonnêteté, puisqu’il s’agit principalement de dissimulation d’éléments et de simulation en vue d’être dédouané·e de l’obligation d’emploi. Il n’est donc pas surprenant que les personnes interrogées expriment la honte qu’elles ressentent d’être soupçonnées d’appartenir à pareil groupe, d’être en bref accusées de ne pas vouloir prendre un emploi. 233 Tout comme les personnes interviewées par Chase et Walker (2013), celles que nous avons rencontrées font le lien entre leur sentiment de honte et les pratiques d’humiliation dont elles ont été victimes dans différents contextes sociaux. Comme l’explique Yann (rentier AI, 24 ans), « on est tout de suite taxés de profiteurs et tout, de sangsues de la société ». Pour Gabriel (chargé de communication, 41 ans), « ça n’est pas très bien vu. Si vous dites ‹ AI ›, […] le soupçon d’abus […] est assez fort… ». Antonio (maître socioprofes- sionnel, 36 ans) abonde aussi dans ce sens : « C’est vite les préjugés, et c’est vite des raccourcis : ‹ Ah, tu es à l’AI ? C’est que tu ne veux rien faire ! › » Aline (auxiliaire de crèche, 32 ans) explique avoir été « catégorisée de fainéante, de personne qui n’a pas envie de travailler », Marie (en attente de décision de l’AI, 48 ans) que « les gens disent tous que […] dès que vous êtes en demande AI, vous êtes un profiteur ». Le processus d’humiliation passe notamment par une mise en doute du caractère invalidant des difficultés de santé qu’elles expérimentent. Mario (en formation de chef de chantier, 56 ans) relate son vécu en ces termes : « Les gens, ils ne vous croient pas. C’est tout de suite, ‹ celui-là qui veut se mettre à l’AI, il ne veut rien faire ! › […] Même ma copine […] ne me croyait pas. » Sonia (rentière AI, 46 ans) raconte à ce propos : « Pour le peu de gens aux- quels je l’ai dit, ils avaient tous la remarque : ‹ […] Ça ne se voit pas que tu es malade. › Je dis : ‹ Pourquoi ? Je dois être amputée des deux jambes pour qu’on me… ? › » Ces discours vont parfois jusqu’à questionner l’existence même de la maladie, notamment en cas de diagnostic psychique. Pour certaines des personnes interviewées, cela revient à leur attribuer la responsabilité de la maladie : « Quand on est atteint d’une maladie psychique, c’est tout juste si l’on n’est pas coupable de ce qui nous est arrivé » (Dominique, libraire, 43 ans). Le processus de l’AI, qui repose sur une capacité toujours présumée, est du même ordre : une personne doit prouver que l’atteinte de santé dont elle souffre est invalidante, et même le jugement de son médecin concernant sa maladie peut être remis en question par les expert·e·s de l’AI (Ferreira 2015). En réaction sans doute à ce jugement social, les personnes que nous avons rencontrées ont souvent tenu à souligner devant nous la réalité de l’atteinte à la santé qu’elles subissent tout en ayant expérimenté la relative inefficacité de ce discours face au regard porté sur elles. Lors des entretiens, ces personnes ont également utilisé deux stratégies de présentation de soi. Première stratégie, signifier qu’elles ont conscience de la solidarité dont elles sont bénéficiaires. Les personnes disent se sentir « redevables » (Laurent, en formation de vendeur, 45  ans) ou avoir « une dette » (Pierre, éducateur, 50 ans). La posture – humble et reconnaissante – correspond à l’image historique du « bon pauvre » (Castel 1995 ; Geremek 234 1978) et permet de se reconstruire sous l’angle de la redevabilité comme sujet conscient qui pense sa position dans le monde. Seconde stratégie, tenter de se démarquer du groupe des « profiteurs » en participant soi-même à nourrir le discours stigmatisant de l’abus. C’est par exemple le cas de Jérôme (en formation de conseiller de vente, 46 ans) : Si les gens ont besoin, s’ils ont vraiment un problème de santé, ça, je suis d’accord, qu’ils ne peuvent plus travailler et puis qu’ils aient une rente de l’AI, ça, je suis d’accord, mais les gens qui peuvent travailler et puis qui se mettent à l’AI, ils essaient toujours d’y entrer, de tou- cher et puis de ne plus rien faire, après ça je ne trouve pas normal. Cette pratique de distinction passe par l’énonciation du caractère invalidant de l’affection dont elles souffrent, comme Florent (employé de banque, 60 ans) qui explique que « sans vouloir profiter, parce que ce n’est pas mon genre du tout, […] c’est une reconnaissance justement de mon incapacité au travail qui est en train de se réaliser ». Dans la vie quotidienne, la honte d’être taxé de profiteur conduit certaines des personnes que nous avons interrogées à des pratiques d’évitement. Elles omettent de parler du fait qu’elles sont en contact avec l’AI. Marie (en attente de décision de l’AI, 48 ans) l’explique : « Je ne dis jamais. […] Parce que les gens disent tous que […] dès que vous êtes en demande AI, vous êtes un profiteur. » Carlos (mécanicien, 26 ans) explique également qu’il évite de dire qu’il est à l’AI. « On me posait souvent des questions : ‹ Oui mais tu fais quoi ? Tu cherches quoi ? › Et souvent je mentais, je disais : ‹ Je cherche du travail. › Mais je savais de base que non, je ne pouvais pas chercher du travail. » Gabriel (chargé de communication, 41 ans) explique : J’étais toujours sous contrat avec mon employeur, […] je pouvais un peu ruser, dire que j’étais encore sous contrat. […] Le côté AI était plus facile à dissimuler, on va dire. En même temps, c’est assez dur de dire… dissimuler. Mais le statut de pensionné AI n’est pas seulement honteux à cause du dis- cours sur l’abus, il l’est également à cause des normes sociales capacitistes (Campbell 2009 ; Goodley 2014) des sociétés contemporaines. La crainte d’être jugé comme handicapé·e, perçu·e comme malade, ou incapable, se conjugue à celle d’être taxé·e de profiteur. Aline (auxiliaire de crèche, 32 ans) ou Martine (infirmière, 47 ans) expliquent par exemple ne pas dévoiler le fait qu’elles ont reçu des prestations de l’AI lorsqu’elles cherchent un emploi adapté à leur santé de peur de ne pas être jugées « compétentes ». 235 Les personnes qui nous disent avoir appris à « accepter » leur statut et ne plus ressentir de honte sont celles qui réussissent à justifier le fait d’être en contact avec l’AI, ce qui signifie – au moins pour un temps – qu’elles ont accepté la position sur l’échelle d’in/validité que leur a assignée l’AI. Brigitte (stagiaire dans un secrétariat, 55 ans) l’exprime en ces termes : Je ne dis pas que je suis contente, du tout. Mais je me suis dit : « Mais, en fait, c’est une assurance sociale, j’ai cotisé pour ça. […] C’est un handicap que j’ai, c’est une fragilité. […] » Il faut apprendre à accepter nos limites. Aline (auxiliaire de crèche, 32 ans) explique qu’elle a accepté « de se dire : ‹ OK, je ne peux plus. […] › Le système est fait pour. Je répète aux gens que le système suisse est fait de filets sociaux, etc., à moi aussi d’accepter ça. » Ce processus d’acceptation est toutefois pénible, comme l’explique Jean- Michel (contrôleur de qualité, 56 ans) : « Au début c’était difficile […]. C’était un travail sur moi-même. […] admettre qu’on peut être fatigué, que les autres ne nous voient plus de la même manière. » Quelques personnes, comme Paul (travailleur social, 33 ans), tiennent un discours plus provocateur qui atteste également la résistance à accepter ce statut : « Moi maintenant je le dis avec fierté. ‹ […] Je sors de l’AI et alors ? › » Laurent (en formation de vendeur, 45 ans) fait de même : « J’en parle […] très facilement. […] Je n’en ai rien à foutre. C’est ma situation actuelle. » L’émotion est « une manière par laquelle nous négocions notre rapport à nous-mêmes, au monde, et aux autres », comme le relève Despret (2005, 283) et le discours sur la honte des personnes que nous avons rencontrées l’illustre bien. Nos données éclairent également quelques-uns des ressorts de la fabrique sociale de l’infériorisation, notamment le fait qu’elle provoque de la résistance. L’évocation de la honte n’est dès lors pas seulement la marque de l’intériorisation de sa place sociale, pour reprendre les termes de Pierre Bourdieu cités plus haut. Le registre de la honte peut aussi être mobilisé afin de contrer la dégradation de son statut social. Cette mobilisation reste néanmoins ambivalente du fait qu’il est impossible en société capacitiste de présenter positivement la situation d’invalide. 10.6 Troisième dimension de la honte : l’effet-miroir Lorsqu’elles se trouvent en réadaptation AI dans le cadre d’une réorientation professionnelle (stage, nouvelle formation) ou pour préparer cette réorien- tation via des programmes de réinsertion, quinze des personnes que nous 236 avons rencontrées évoquent une troisième dimension du registre la honte que nous nommons, dans une perspective quelque peu différente de l’inte- ractionnisme symbolique de Charles Horton Cooley (1902), « effet-miroir ». En effet, dans le sens où nous l’utilisons, cet effet-miroir découle du fait que dans les programmes de réadaptation, d’une part les individus se trouvent dans une structure collective qui les met en présence d’autres personnes, ce qui leur donne un reflet de leur situation, et d’autre part que la conception ou les activités organisées leur renvoient une image de leur condition présente. La technologie de rapprochement de l’emploi via ces programmes leur pose différents problèmes. D’abord, dans les programmes de réadaptation, les personnes que nous avons rencontrées ont côtoyé différents individus considérés par les politiques sociales comme ayant besoin d’une réadaptation professionnelle. Non seu- lement d’autres assuré·e·s de l’AI (dont l’invalidité peut découler de toute autre cause), mais encore des personnes à l’assistance sociale, ou ayant un problème d’addiction, ou dépendant de la législation sur l’asile, etc. Pour garantir leur viabilité économique, les prestataires de programmes ont décidé en effet d’accueillir des publics divers dont la caractéristique commune est l’éloignement de l’emploi. Voir d’autres personnes dans les mêmes programmes, comme en miroir, donne une expression concrète à leur sentiment de s’écarter des normes capacitistes. Or, les personnes que nous avons interrogées jugent fréquemment leur situation incommensurable à celles des autres participant·e·s, et elles cherchent dès lors à s’en distinguer. Laurent (en formation de vendeur, 45 ans) rapporte qu’il a passé du temps dans une structure qui était « pour les toxicomanes et tout ça. Mais, moi, j’ai été là-bas, c’était aussi pour me poser et puis parce que j’avais un médicament qui était très fort après le burn-out […] ce n’était pas évident. Là, on retombe assez bas ». Carlos (mécanicien, 26 ans) explique : « Les ateliers protégés qu’ils m’ont proposés, […] c’étaient des gens encore plus en difficulté que moi, c’étaient des gens en chaise roulante. […] Pourquoi on veut me mettre là-dedans ? » Marie (en attente de décision de l’AI, 48 ans) relate : « Je me suis retrouvée avec des gens, des requérants et des gens au chômage, qui n’en avaient rien à foutre d’être là. » L’effet-miroir provoque de la honte, mais aussi de la résistance en vue de contrecarrer les regards stigmatisants sur leur invalidité : tout en étant elles-mêmes stigmatisées, elles essaient de se distancier d’autres publics stigmatisés. Ensuite, cet effet-miroir peut également être lié à la conception même du programme, qui peut donner une impression de retour à une étape antérieure de sa vie, à l’instar de Mario (en formation de chef de chantier, 56 ans) qui trouve difficile de « se remettre sur les bancs de l’école avec des gens de 19 ans ». 237 Marie (en attente de décision de l’AI, 48 ans) raconte : « Vous vous rendez compte ? J’ai dû faire des tests de calcul écrit. Comme les petits jeunes, on m’a fait faire de la grammaire. On m’a fait faire du calcul aussi. » Sonia (rentière AI, 46 ans) s’insurge également contre cette situation : « Vous ne pouvez pas demander à une personne […] de mon âge, à 45 ans, ayant en tout cas 20 ans d’expérience professionnelle, de retourner à l’école scolaire, faire un bond en arrière de 20 ans. » Ces personnes expérimentent une forme de régression dans leur parcours de vie, un déclassement dans la hiérarchie des âges. Le « bond en arrière » dont parle Sonia n’est en effet pas seulement tem- porel, il est également statutaire. C’est le retour à un statut de « personne en formation ». Il faut ici se souvenir que l’âge n’est pas une donnée naturelle ou un simple outil de classification, mais un rapport social (Jenny 1995 ; Perriard 2017 ; Tabin et Perriard 2014). Aux catégories d’âges instituées par l’État correspondent des normes sociales qui régissent notamment le rapport à l’emploi qui est la norme entre 25 et 64 ans en Suisse. Dans un parcours de vie standard, constitué par la séquence enfance – formation – emploi/ famille – retraite, « la position dominante est occupée par la personne en emploi […], les autres positions sociales étant pensées à partir de cette position dominante : c’est le but à atteindre plus tard pour les enfants et les jeunes, c’est le fait de l’avoir occupée qui justifie les prestations de retraite » (Tabin et Perriard 2014). Et c’est cette position dominante de travailleur adulte qui confère statut, autorité, argent et (relative) autonomie. Ce statut est remis en question par le processus de réadaptation qui repose principalement sur des formations et sur des stages. Le statut inférieur lié à cette étape est confirmé dans les interactions avec l’AI. Brigitte (stagiaire dans un secrétariat, 55 ans) l’explique clairement : Ce qui a été un peu difficile aussi, c’était de se dire qu’on est stagiaire à 55 ans après avoir travaillé pendant 38 ans… Stagiaire, et de ne pas aller seule à un entretien d’embauche. […] C’est une coach [de l’AI] qui nous accompagne, ce qui est très bien. Mais il y a le statut où l’on travaille et puis là, c’est un autre statut. Des attributs considérés comme normaux à l’âge adulte, tels l’autonomie, sont remis en question, ce qui est qualifié par Julie (employée de commerce, 27 ans) d’ « infantilisation », avec des agent·e·s qui évaluent l’activité réalisée en disant : « ‹ Ça, c’est bien. Ça, c’est moins bien. Faudrait faire ci et ça. › Je ne l’ai pas très bien vécu. […] C’est vraiment cet aspect un peu infantilisant des fois… » Dans son cas, elle a dû remplir une évaluation dont l’une des questions portait sur « la présentation et l’hygiène corporelle ». Le sentiment 238 de honte de Julie est lié ici au fait d’être évaluée sur des compétences qui sont de l’ordre de l’éducation. Or, ce sont les enfants que l’on éduque. Enfin, l’effet-miroir peut être lié aux activités que les personnes effectuent dans le cadre des programmes de réadaptation. Les personnes que nous avons interrogées évoquent par exemple des tâches telles que « trier des habits, vous imaginez l’estime de soi après tout ça ? » (Marie, en attente de décision de l’AI, 48 ans), ou des activités considérées comme vides de sens et sans utilité : « C’était vraiment du temps perdu. […] C’était une semaine où vous restez devant l’ordinateur pour apprendre un petit peu à le manier » (Mario, en formation de chef de chantier, 56 ans), ou encore comme inadaptées : « J’étais dans un truc pour faire conciergerie. […] Je ne voulais pas devenir concierge […] mais… […] assistant en stérilisation » (Jean, rentier AI, 48 ans). Dans les mécanismes de la réadaptation, les personnes interviewées expéri- mentent donc diverses formes de déviation par rapport à la « normalité », qu’il s’agisse d’un décalage avec les autres personnes participant aux programmes, par rapport à l’âge « normal » de la formation, ou de la spécificité du type de tâches qui leur sont proposées. La mobilisation du registre de la honte fonc- tionne, dans le cadre de cette troisième dimension, comme une manière de se reconstruire comme sujet en se distanciant des catégories et des activités dans lesquelles l’AI les place. Cette mobilisation peut-elle subvertir les schémas classificatoires ? Lorsque les personnes interrogées expriment leur critique des actions de l’AI et leur résistance aux catégorisations faites par les agent·e·s de l’AI, c’est sans doute partiellement le cas, car elles se présentent comme des sujets et non comme des objets d’une action publique. Mais cette mobilisation signifie également qu’elles ont intériorisé l’idéologie capacitiste (Campbell 2008 ; Siebers 2013) car elles tentent de se distinguer des autres en les définissant comme « plus handicapées », « plus malades » ou comme appartenant à des groupes sociaux subordonnés (personnes réfugiées, au chômage…), ou des activités qui sont proposées en les présentant comme inadaptées à leur situation, voire indignes. 10.7 Conclusion Notre analyse a montré que les personnes dont la capacité de travail est remise en cause par une atteinte à la santé se réfèrent spontanément à la honte, confir- mant les conclusions d’études scientifiques antérieures qui soulignaient le rôle joué par la honte dans la reproduction des hiérarchies sociales. Nos résultats confirment la valeur sociale centrale de la capacité de travail et montrent que la honte fonctionne comme une force qui maintient les divisions sociales. 239 Une analyse plus fine indique toutefois que ses tonalités sont diverses. La honte se réfère dans certains cas au nouveau positionnement social que les personnes expérimentent parce qu’elles sont éloignées de l’emploi et en contact avec l’AI, une institution qui entérine cette distance. La honte découle de la hiérarchie capacitiste qui oblige les personnes à incorporer un statut infériorisé, celui d’invalide ou de personne en risque d’invalidité, pour avoir droit aux prestations. Dans d’autres cas, la honte vient de la crainte ou de l’expérience d’un jugement social. Comme le discours moral sur l’abus tend à généraliser le soupçon sur les personnes invalides, la honte procède dans ce cas d’un processus de stigmatisation et d’humiliation. Enfin, la honte peut survenir comme conséquence de l’effet-miroir lié à la participation à des programmes de réadaptation. Elle renvoie alors à un décalage par rapport à la normalité (âge standard de la formation, confrontation avec d’autres per- sonnes déclarées inaptes à l’emploi, type de tâches à effectuer) et concrétise l’éloignement de la norme dominante de capacités, l’emploi. Toutefois, la honte n’est pas seulement une émotion négative dont la fierté serait l’antonyme. L’analyse des références à la honte que nous avons effectuée met au jour des usages plus complexes de cette émotion qui permettent de s’associer au groupe dominant des personnes en capacité d’emploi, de rede- venir sujet. Il ne s’agit pas d’une remise en question frontale de la hiérarchie capacitiste, mais d’un déplacement des significations attachées à l’invalidité à l’aide de stratégies d’évitement, de démarcation, ou encore en mettant en évidence la violence des réassignations identitaires découlant du contact avec l’AI. La mobilisation du registre de la honte permet dès lors plutôt de mettre en question, à un niveau individuel, les catégories auxquelles sont assignées les personnes concernées : ces personnes sont des sujets, elles ne sont pas réductibles au statut d’invalide (permanent ou en risque d’invalidité) que leur confère le contact avec l’AI. La honte s’inscrit alors dans le double registre des émotions décrit par Vinciane Despret (2005, 244), « de ce que nous fabriquons et de ce qui nous fabrique ». La honte, parmi d’autres émotions, « particip[e] alors activement à la création du social » (Despret 2005, 246), elle peut être une force pour constituer autrement la « normalité ». Si les personnes que nous avons rencontrées s’engagent surtout dans des négociations individuelles de leur identité en mobilisant le registre de la honte, des manifestations d’action collective apparaissent également en Suisse, comme les mouvements Disability Pride5 qui pourraient contribuer à subvertir de 5 En Suisse alémanique : http://disability-pride.ch/fotos-umzug-2017; en Suisse romande, une première Mad Pride a eu lieu en 2019 à Genève, la deuxième édition étant prévue en 2021 à Berne http://www.coraasp.ch/pages/communication/journee-mondiale-de-la-sante- mentale/mad-pride-defilons-pour-la-diversite-le-101019-a-geneve-331 (4.3.2021). 240 manière moins individualisante la hiérarchie capacitiste et permettre de penser autrement le futur de l’emploi des personnes catégorisées comme invalides. 10.8 Références bibliographiques Ahmed, Sara. 2004. The Cultural Politics of Emotion. Edinburgh: Edinburgh University Press. Baker, Helen. 2013. The Significance of Shame in the Lives of Women who Ex- perience Male Violence. Liverpool Law Review, 34: 145‒171. Barrett, Timothy. 2016. Bourdieu, Hysteresis, and Shame: Spinal Cord Injury and the Gendered Habitus. Men and Masculinities, 21(1): 35‒55. Biddle, Jennifer. 1997. Shame. Australian Feminist Studies, 12(26): 227‒239. Bourdieu, Pierre. 1997. Méditations pascaliennes. Paris: Le Seuil. Bourdieu, Pierre. 1998. La domination masculine. Paris: Le Seuil. Bourdieu, Pierre. 2012. Sur l’État. Cours au Collège de France (1989‒1992). Paris: Raisons d’agir/Seuil. Bourdieu, Pierre et Yvette Delsaut. 1975. Le couturier et sa griffe : contribution à une théorie de la magie. Actes de la Recherche en Sciences Sociales, 1(1): 7‒36. Campbell, Fiona Kumari. 2008. Exploring Internalized Ableism using Critical Race Theory. Disability and Society, 23(2): 151‒162. Campbell, Fiona Kumari. 2009. Contours of Ableism. The Production of Disability and Abledness. Basingstoke and New York: Palgrave Macmillan. Castel, Robert. 1995. Les métamorphoses de la question sociale : une chronique du salariat. Paris: Fayard. Chase, Elaine et Robert Walker. 2013. The Co‒construction of Shame in the Con- text of Poverty: Beyond a threat to the Social Bond. Sociology, 47(4): 739‒754. Clough, Patricia T. et Jean Halley (dir.) 2007. The Affective Turn: Theorizing the Social. Durham: Duke University Press. Conseil fédéral. 2017. Révision de la loi fédérale sur la partie générale du droit des assurances sociales (LPGA). Rapport explicatif relatif à l’ouverture de la procédure de consultation. Berne, https://www.bsv.admin.ch/dam/bsv/fr/dokumente/the menuebergreifend/vernehmlassungen/vernehmlassung-revision-atsg-bericht. pdf.download.pdf/vernehmlassung-revision-atsg-bericht-fr.pdf (4.3.2021) Cooley, Charles Horton. 1902. Human Nature and the Social Order. New York: Routledge. Creed, Douglas, Ashley Hudson, Gerardo A.Okhuysen, et Kristin Smith-Crowe. 2014. Swimming in a sea of Shame: Incorporating Emotion into Explanations of Institutional Reproduction and Change. Academy of Management Review, 39(3): 275‒301. 241 Davis, Lennard J. 1995. Enforcing Normalcy: Disability, Deafness, and the Body. London: Verso. De Gaulejac, Vincent. 1996. Les sources de la honte. Paris: Desclée de Brouwer. Despland, Béatrice. 2012. L’obligation de diminuer le dommage en cas d’atteinte à la santé. Son application aux prestations en espèces dans l’assurance maladie et l’assurance invalidité. Analyse sous l’angle du droit d’être entendu. Genève- Zurich-Bâle: Schulthess. Despret, Vinciane. 2005. Ces émotions qui nous fabriquent. Ethnopsychologie de l’authenticité. Paris: Les Empêcheurs de penser en rond/Le Seuil. Driessens, Kristel. 2010. La relation entre intervenants sociaux et personnes vi- vant dans la pauvreté. Affronter l’exclusion et la dépendance. Pensée Plurielle, 25(3): 91‒105. Ferreira, Cristina. 2015. Invalides psychiques. Experts et litiges. Lausanne: Antipodes. Ferreira, Cristina et Arnaud Frauenfelder. 2007. « Y en a qui abusent…». Identifier, gérer et expertiser des ayants droit de la politique sociale. Carnets de Bord en Sciences Humaines, 13: 3‒6. Fox, Catherine Olive-Marie. 2014. Toward a Queerly Classed Analysis of Shame: Attunement to Bodies in English Studies. College English, 76(4): 337‒356. Fracheboud, Virginie. 2015. L’introduction de l’assurance invalidité en Suisse (1944‒1960). Tensions au cœur de l’État social. Lausanne: Antipodes. Fraser, Nancy et Linda Gordon. 1994. A Genealogy of Dependency: Tracing a Keyword of the U.S. Welfare State. Signs, 19(2): 309‒336. Frauenfelder, Arnaud et Carola Togni. 2007. Les « abus » à l’aide sociale : une rhétorique au cœur du gouvernement de la misère. Carnets de Bord en Sciences Humaines, 13: 43‒55. Fullagar, Simone. 2003. Wasted lives. The Social Dynamics of Shame and Youth Suicide. Journal of Sociology, 39(3): 291‒307. Geremek, Bronislaw. 1978. La potence ou la pitié. L’Europe et les pauvres du Moyen- Âge à nos jours. Paris: Gallimard. Glaser, Barney G. et Anselm A. Strauss. 2010. La découverte de la théorie ancrée. Stratégies pour la recherche qualitative. Paris: Armand Colin. Goffman, Erwin. 1973. La mise en scène de la vie quotidienne. Tome II. Les relations en public. Paris: Minuit. Goodley, Dan. 2014. Dis/ability Studies: Theorising Disablism and Ableism. Lon- don: Routledge. Goodley, Dan, Rebecca Lawthom, et Katherine Runswick-Cole. 2014. Dis/ ability and Austerity: Beyond Work and Slow Death. Disability and Society, 29(6): 980‒984. Holden, Chris. 2003. Decommodification and the Workfare State. Political Studies Review, 1(3): 303‒316. Hughes, Bill. 2015. Disabled People as Counterfeit Citizens: The Politics of Re- sentment Past and Present. Disability and Society, 30(7): 991‒1004. 242 Jenny, Jacques. 1995. Rapports sociaux de sexe et autres rapports de dominance sociale : pour une intégration conceptuelle des rapports sociaux fondamentaux. Cahiers du GEDISST, 13: 109‒130. Kosofsky Sedgwick, Eve. 1993. Queer Performativity: Henry James’s the Art of the Novel. GLQ : A Journal of Lesbian and Gay Studies, 1(1): 1‒16. Lordon, Frédéric. 2013. La société des affects. Pour un structuralisme des passions. Paris: Le Seuil. Loveday, Vik. 2016. Embodying Deficiency through ‹ Affective Practice › : Shame, Relationality, and the Lived Experience of Social Class and Gender in Higher Education. Sociology, 50(6): 1140‒1155. Mead, George H. 1963. L’esprit, le soi et la société. Paris: Presses universitaires de France. Mitchell, David T. et Sharon L. Snyder. 2000. The Body and Physical Difference. Discourses of Disability. Ann Arbor: University of Michigan Press. Moore, Allison. 2016. Shame on You: the Role of Shame, Disgust and Humilia- tion in Media Representations of ‹ Gender-Fraud › Cases. Sociological Research Online, 21(2): 1‒18. https://doi.org/DOI:10.5153/sro.3942. Munt, Sally R. 2008. Queer Attachments. The Cultural Politics of Shame. Alder- shot: Ashgate. OECD. 2009. Sickness, Disability and Work: Keeping on Track in the Economic Downturn. Stockholm: Organisation for Economic Co-operation and Devel- opment, https://www.oecd.org/els/emp/42699911.pdf (4.03.2021). Offe, Claus. 1984. Contradictions of the Welfare State. London: Hutchinson. Parker Harris, Sarah, Randall Owen, Karen R. Fisher, et Robert Gould. 2014. Human Rights and Neoliberalism in Australian Welfare to Work Policy: Expe- riences and Perceptions of People with Disabilities and Disability Stakeholders. Disability Studies Quarterly, 34(4). Perriard, Anne. 2017. Les figures de la dépendance problématique des adultes : analyse intersectionnelle des politiques sociales liées à l’âge et à l’emploi dans le canton de Vaud. PhD dissertation, Faculté des Sciences Sociales et Politiques, Université de Lausanne. Piecek, Monika, Jean-Pierre Tabin, Céline Perrin, et Isabelle Probst. 2017. La normalité en société capacitiste. Une étude de cas (assurance invalidité suisse). SociologieS [Online], Theory and research, https://sociologies.revues.org/6412 (4.03.2021) Piecek, Monika, Jean-Pierre Tabin, Céline Perrin, et Isabelle Probst. 2019. The ‹ Compliant ›, the ‹ Pacified › and the ‹ Rebel › : Experiences with Swiss Disability Insurance. Disability and Society, 34(4): 583‒606. Pires, Alvaro. 1997. Échantillonnage et recherche qualitative : essai théorique et méthodologique. In Jean Poupart, Jean-Pierre Deslauriers, Lionel-Henri Groulx, Anne Laperrière, Robert Mayer et Alvaro Pires (dir.), La recherche qualitative. Enjeux épistémologiques et méthodologiques. Montréal: Gaëtan Morin, 113‒169. 243 Probst, Isabelle, Jean-Pierre Tabin, et Nelly Courvoisier. 2015. De la réparation à la réversibilité. Un nouveau paradigme dans l’assurance invalidité ? Revue Suisse de Sociologie, 41(1): 101‒117. Probst, Isabelle, Jean-Pierre Tabin, Monika Piecek-Riondel et Céline Perrin. 2016. L’invalidité : une position dominée. Revue Française des Affaires Sociales, 4: 89‒105. Probyn, Elspeth. 2005. Blush: Faces of Shame. Minneapolis: University of Min- nesota Press. Reeve, Donna. 2012. Psycho-Emotional Disablism: The Missing Link? In Nick Watson, Alan Roulstone, et Carol Thomas (dir.), Routledge Handbook of Dis- ability Studies. New York: Routledge, 78‒92. Scheff, Thomas. 2000. Shame and the Social Bond: A Sociological Theory. Socio- logical Theory, 18(1): 84‒99. Shildrick, Margrit. 2002. Embodying the Monster: Encounters with the Vulnerable Self. London: Sage. Siebers, Tobin. 2013. Disability and the Theory of Complex Embodiment. For Identity Politics in a New Register. In Lennard J. Davis (dir.), The Disability Studies Reader. New York: Routledge, 278‒297. Skeggs, Beverley. 1997. Formations of Class & Gender: Becoming Respectable. London: Sage. Skeggs, Beverley. 2004. Class, Self, Culture. London: Routledge. Skeggs, Beverley. 2009. The Moral Economy of Person Production: The Class Relations of Self-Performance on ‹ Reality › Television. The Sociological Review, 57(4): 626‒644. Stiker, Henri-Jacques. 2005. Corps infirmes et sociétés : essais d’anthropologie his- torique. Paris: Dunod. Tabin, Jean-Pierre, Arnaud Frauenfelder, Carola Togni, et Véréna Keller. 2010. Temps d’assistance. Le gouvernement des pauvres en Suisse romande depuis la fin du XIXe siècle. Lausanne: Antipodes. Tabin, Jean-Pierre et Anne Perriard. 2014. Le rapport social d’âge dans les poli- tiques sociales. Revue ¿ Interrogations ?, 19, www.revue-interrogations.org/ Le-rapport-social-d-age-dans-les (4.03.2021). Walker, Robert. 2014. The Shame of Poverty. Oxford: Oxford University Press. Weiss, Karen G. 2010. Too Ashamed to Report: Deconstructing the Shame of Sexual Victimization. Feminist Criminology, 5(3): 286‒310. Yates, Scott et Alan Roulstone. 2013. Social Policy and Transitions to Training and Work for Disabled Young People in the United Kingdom: Neo-Liberalism for Better and for Worse? Disability and Society, 28(4): 456‒470. Young, Iris Marion. 2003. Autonomy, Welfare Reform and Meaningful Work. In Eva Feder Kittay et Ellen K. Feder (dir.), The Subject of Care. Feminist Per- spectives on Dependency. Lanham: Rowman Littlefield, 40‒60. Part V The Future of Work for Academics 247 11 Partir un jour, sans retour ? Injonctions et pratiques de la mobilité internationale dans l’espace académique suisse contemporain – le cas des postdocs Pierre Bataille et Marie Sautier1 11.1 Introduction « Faire un postdoc », « partir à l’étranger », « publier », comptent au nombre des multiples injonctions auxquelles les jeunes docteur·e·s en quête d’un poste académique stable sont encouragé·e·s à se conformer. Porté par de nombreuses institutions académiques européennes, ce discours intervient dans un contexte marqué par la forte augmentation du nombre de postdoctorant·e·s depuis le début des années 1990 (Theodosiou et al. 2012), et une internationalisation croissante du champ scientifique et des carrières des personnels académiques (Goastellec 2017). Les données analysées dans cette enquête rassemblent des questionnaires et entretiens réalisés auprès de postdoctorant·e·s ayant travaillé dans une même université suisse au cours des années 2010. Nous montrons que si l’expérience postdoctorale à l’étranger tend à s’imposer comme une des conditions sine qua non pour accéder à un poste académique stable dans le contexte hel- 1 Ce travail a fait l’objet d’une publication initiale dans la revue Philosophia Scientiae en 2020 : Pierre Bataille et Marie Sautier (2020). Ce « qu’être postdoc » veut dire. Cheminements postdoctoraux en Suisse, circa 2010. In Philosophia Scientiae, 24(1) ; https://doi.org/10.4000/philosophiascientiae.2207. Le chapitre présent est une repu- blication dans une forme légèrement modifiée, effectuée à la demande des éditeurs/ rices de cet ouvrage. Nous remercions Philosophia Scientiae pour son accord. Cette recherche a été financée par le Fonds National Suisse et le NCCR LIVES (grant number: 51NF40-185901). Les auteur·e·s remercient le Fonds National Suisse pour son soutien financier. 248 vétique, ce modèle de carrière s’hybride avec des formes plus anciennes de structuration des carrières académiques, en lien avec l’histoire particulière du système universitaire suisse. Notre contribution est organisée comme suit : nous contextualisons ques- tions de recherche dans un premier temps en rappelant les mutations récentes de l’emploi académique et la manière dont elles ont été interprétées dans la littérature contemporaine. Nous présentons ensuite les particularités de l’emploi académique suisse et détaillons le modèle analytique mobilisé pour rendre compte des logiques guidant les parcours et expériences des postdocs que nous avons rencontré·e·s. Nous proposons enfin une analyse de l’espace des carrières caractérisant nos enquêté·e·s. Nous observerons qu’au-delà des marqueurs d’excellence mis en avant par les instances évaluatrices – parmi lesquelles le nombre de publications dans des revues à comité de lecture ou encore la mobilité internationale – certains éléments de parcours comme la socialisation conjugale participent fortement à orienter les carrières. Enfin, nous verrons comment ces éléments objectifs de structuration des carrières permettent de rendre compte et d’éclairer la pluralité des expériences qui se font jour dans les entretiens que nous avons menés. 11.2 Vers une uniformisation des débuts de la carrière académique ? Depuis une vingtaine d’années, le nombre de chercheur·e·s employé·e·s sur des contrats « postdoctoraux » a connu une augmentation très rapide dans la plupart des universités européennes et nord-américaines. S’il est peut-être moins nouveau qu’il n’y paraît2, ce phénomène dit de la « bulle postdoctorale » a attiré l’attention des pouvoirs publics et des instances professionnelles, en particulier dans les sciences de la vie et plus généralement dans les sciences dites « dures » où le phénomène est particulièrement développé (Powell 2015 ; Theodosiou et al. 2012). Cette multiplication des contrats postdoctoraux s’inscrit plus généralement dans un mouvement de « dé-standardisation » des carrières académiques (Enders et Musselin 2008) : dans la plupart des espaces nationaux, on observe en effet une relative stagnation – voire une diminu- 2 L’histoire des universités allemandes est ainsi émaillée de crises dues à une sur- représentation de personnels académiques employés sur des postes non-stables, comme le rappellent Charle et Verger (2012, 120‒122) à travers le cas du mouvement des Nicht- ordinarien à la veille de la première guerre mondiale. Le degré d’internationalisation de la recherche ne suit pas une évolution linéaire et connaît d’importantes variations et inversions de tendances depuis la création des premières universités, comme le souligne notamment Gingras (2002). 249 tion – du nombre de postes académiques stables alors que se multiplient les postes non-stables. De fait, pour une part grandissante du personnel acadé- mique, ce qui représentait jusqu’à présent un état temporaire de la carrière, c’est à dire l’engagement sur des postes non-stables entre la fin du doctorat et le recrutement sur des postes stables ou tenure track, tend à devenir une modalité d’emploi durable, ainsi que le montre, par exemple, la « tripartition » du salariat académique en Grande-Bretagne (Paye 2015). Les travaux des quinze dernières années portant sur les parcours, pratiques et représentations des postdocs, que ce soit en sociologie ou en sciences de l’éducation, témoignent d’un intérêt croissant pour cette fraction encore peu connue du corps académique contemporain. Une part importante de des travaux relaie ainsi les difficultés que rencontrent ces « travailleurs invisibles » (Can- twell et Lee 2010) évoluant dans les entreprises du « capitalisme académique » (Ylijoki 2003) contemporain, en particulier les universités et programmes de recherches internationaux ; les mécanismes de compétition et la pression qu’ils suscitent favorisent des conflits entre sphères familiales, personnelles et professionnelles, qui affectent particulièrement les trajectoires féminines (del Río Carral et Fusulier 2013 ; Dorenkamp et Süß 2017 ; Fusulier et del Río Carral 2012 ; Henderson 2020). Ces travaux ont largement contribué à nuancer l’image romantisée de « créatifs » indépendants mus par le seul amour de la science, régulièrement relayée dans les médias (Armano et Murgia 2013 ; Murgia et Poggio 2014). Les recherches les plus récentes pointent d’ailleurs que ceux et celles qui bifurquent hors de la carrière académique, par exemple dans l’industrie ou l’administration, tendent à trouver des conditions plus favorables à leur qualité de vie personnelle et professionnelle, avec d’impor- tantes variations selon les contextes nationaux cependant (Bataille et al. 2017 ; Dorenkamp et Weiß 2017 ; Fusulier et Dubois-Shaik 2016). Si certaines analyses longitudinales montrent que « faire un postdoc », en particulier à l’étranger, a bien un impact sur les chances de décrocher un poste académique stable, les effets à long terme du postdoctorat sur la progression de carrière, le niveau de salaire, ou encore la productivité scientifique, sont très loin d’être évidents (Jöns 2011 ; Yang et Webber 2015). Effectuer un postdoctorat à l’étranger3, ou collaborer à l’étranger4, est davantage identifié 3 Pour une très grande partie des postdocs en provenance d’Europe ou d’Amérique du Nord, la mobilité internationale est avant tout envisagée comme une étape transitoire, avant le retour au pays d’origine (Cantwell 2011). On retrouve des motivations similaires chez les cadres internationaux, pour qui la mobilité représente avant tout une ressource à mobiliser dans le cadre de leur espace national d’origine (Colombi 2016). 4 Zippel (2017) pointe par exemple l’ambivalence des discours entourant les pratiques de collaborations internationales aux US. 250 en Europe continentale que dans le monde anglosaxon comme un atout (Ortlieb et Weiss 2018) ou un passage obligé (Sautier 2021) par les jeunes chercheur·e·s en quête d’un poste stable. Nombres de dispositions politiques promouvant, parfois de manière coer- citive, la mobilité des chercheur·e·s, prennent appui sur des argumentaires tels que ceux développés dans le cadre du European Research Era (ERA), et associant mobilité internationale, qualité de la recherche et compétitivité, dans la grande « course » mondiale des universités (Musselin 2017). L’internationa- lisation dès les premières étapes de la carrière académique semble largement bénéfique pour les institutions d’enseignement supérieur tant du point de vue économique que réputationnel (Horta 2009). D’autres travaux indiquent cependant que le lien entre mobilité individuelle et internationalisation de la production scientifique et intellectuelle reste relativement ténu (Ackers 2008), et que certaines stratégies de « mobilités virtuelles » (rendues possibles grâce au développement du Web notamment) ou de mobilités courtes (Bernela et Milard 2016) peuvent s’avérer tout autant, ou davantage intéressantes en termes économiques, écologiques et intellectuels (Storme et al. 2017). L’injonction à la mobilité participe en outre à décupler la vulnérabilité des jeunes docteur·e·s, surtout lorsqu’ils et elles viennent de pays des Suds globalisés – envisageant plus rarement un retour dans l’espace national où ils et elles ont effectué leur formation doctorale et bénéficiant de fait d’une marge de manœuvre souvent étroite pour négocier leurs conditions de travail (Cantwell 2011 ; Shinozaki 2017). Elle contribue également à renforcer le formatage genré des carrières (Fusulier et Dubois-Shaik 2016 ; Jöns 2011), même si les séjours à l’étranger conduisent sous certaines conditions à des couples d’hommes ou de femmes chercheur·e·s à expérimenter des formes d’arrangements conjugaux moins traditionnels (Schaer et al. 2017). Les constats d’une dé-standardisation des carrières académiques dans les contextes nationaux, d’une internationalisation du marché du travail et les effets combinés de ces deux tendances sur les carrières et représentations des jeunes chercheur·e·s sont aujourd’hui relativement bien connus (Fumasoli et al. 2015). S’ils correspondent à une tendance de fond bien documentée, les travaux cités tendent néanmoins à donner une image relativement homogène des évolutions récentes de l’emploi académique. Premièrement, ils prennent assez peu en compte les possibles reconfigurations spécifiques des modèles nationaux de carrière. Ensuite, fondés principalement sur l’analyse des carrières de spécialistes des sciences expérimentales dites « dures », ils tendent également à sous-estimer le poids des différences disciplinaires dans la structuration des parcours professionnels. Enfin, s’appuyant soit sur des analyses quantitatives permettant d’identifier les caractéristiques individuelles (dé)favorisant l’accès 251 à un poste stable), soit sur des analyses qualitatives analysant le sens que les postdocs attribuent à leurs pratiques professionnelles, en particulier ceux et celles situé·e·s dans des postes précaires, ces travaux produisent, dans un cas comme dans l’autre, une vision parcellaire de la réalité objective et subjective du vécu des postdocs. L’objectif des analyses développées dans cet chapitre est double. Il s’agit d’une part d’avancer des arguments en faveur d’une inclusion plus systématique des contextes nationaux dans la compréhension des carrières et des expériences de mobilité des postdoctorant·e·s en Suisse – et au-delà. Il s’agit d’autre part de questionner comment ces métamorphoses peuvent être sensibles à des variations disciplinaires. Ce chapitre ambitionne ainsi de montrer que, loin de remplacer totalement les modalités nationales de recrutement pour des postes académiques stables, on observe une hybridation entre les tendances globales de dé-standardisation et d’internationalisation des carrières et les modes tradi- tionnels de recrutement. En lien avec cette préoccupation et dans la continuité de certains travaux critiques quant à la promotion de la mobilité comme seul critère explicatif de stabilisation des carrières académiques (Goastellec 2016), trois questions de recherche orienteront principalement nos analyses : dans quelle mesure avoir « fait un postdoc » est-il un atout valable dans toutes les disciplines et les espaces académiques nationaux ? Dans quelle mesure d’autres ingrédients des parcours individuels, notamment la socialisation conjugale, informent-ils également les parcours postdoctoraux et les chances d’accéder à un emploi académique pérenne ? Enfin, comment ces différents éléments imprègnent-ils l’expérience vécue des postdoctorant·e·s et l’image qu’ils et elles se font du travail et de la carrière académique ? 11.3 Une analyse socialement ancrée des expériences postdoctorales suisses 11.3.1 L’emploi académique contemporain en Suisse : entre internationalisation à marche forcée et fragmentation avancée des carrières Nous nous appuyons ici sur le cas suisse5, qui présente plusieurs particularités pertinentes en relation avec nos différentes questions de recherche. Premiè- 5 Les données analysées ont été recueillies dans le cadre du projet de recherche GARCIA (Gendering Academy and Research : Combating Inequalities and Assymetries, http:// garciaproject.eu/1) et du NCCR LIVES. Les entretiens ont été essentiellement réalisés par P. Bataille et ont été retranscrits en intégralité par la suite par Pralong, Posse et Ksontini (2015). Les deux auteur·e·s tiennent également à remercier les autres 252 rement, après avoir connu une période de relative fermeture, de la première guerre mondiale aux années 1960, l’espace académique suisse a connu une forte internationalisation au cours du dernier tiers du XXe siècle (Rossier et al. 2015), particulièrement accentuée depuis les années 2000 sous l’effet de mesures incitatives ou coercitives visant à placer les universités suisses sur le marché académique mondial (Benninghoff et al. 2009). A l’Université de Lausanne, par exemple, la durée maximale d’embauche sur un ou plusieurs postes académiques non-stables ne peut dépasser cinq années. Mis en place en 2007, ce règlement a notamment pour conséquence de limiter les possibilités d’embauche de postdoctorant·e·s dans le vivier local et favorise un turnover important au sein des équipes de recherche. Les effets des politiques d’inter- nationalisation mises en place au niveau fédérales ou locales sont aujourd’hui très nets : avec un peu plus de 40 pour cent d’étrangers parmi les membres du personnel académique en 2017 – tous statuts confondus –, la Suisse se place en tête des pays européens sur le plan de l’internationalisation des carrières de recherche (EACEA 2017, 103). Figure 1: Evolution de la composition de l’emploi académique en Suisse (1980–2014) 50 Personnels administratifs/techniques 40 Assistant·e·s & collaborateur·rice·s scientifique 30 20 Autre enseignant·e·s 10 Professeur·e·s 1980 1990 2000 2010 Source : Office fédéral de la statistique (2017). Lecture : en 1980, les professeur·e·s représentaient 10 % de la masse salariale du personnel académique des hautes écoles universitaires suisses. En 2010, ce groupe ne représente plus que 5 %. membres de l’équipe GARCIA de l’UNIL (M. Del Rio Carral, F. Fassa, G. Goastellec, S. Kradolfer et N. Le Feuvre) pour l’ensemble du travail en commun effectué au cours de ce projet. Composition (en %) 253 Cette internationalisation à marche forcée s’est accompagnée d’une remise en question du modèle de carrière traditionnel. Le système universitaire suisse, fortement influencé par le modèle allemand malgré des variations cantonales, a longtemps été organisé autour de chaires composées d’un·e professeur·e – généralement engagé·e sur un poste stable – et d’un groupe plus ou moins large d’assistant·e·s et/ou de Privatdozent6, engagé·e·s sur des postes non-stables. Ces dernier·ère·s constituent le principal vivier de recrutement lorsque le ou la professeur·e titulaire arrive à la retraite, ou décide de partir sur un autre poste. Dans ce système dit du « survivor » (Enders et Musselin 2008), les stabilisations interviennent généralement relativement tard dans les carrières et au prix de longues années passées sur des postes précaires. Bien que, dans de telles conditions, l’accès à un poste stable était souvent relativement improbable, le prestige que représentait la fonction et le ratio entre assistant·e·s et professeur·e·s étaient tels que les impétrant·e·s pouvaient s’accrocher à l’espoir d’être un jour titularisé·e·s. Le développement de la recherche « sur projet » (Benninghoff et al. 2009), la mise en place de nouveaux programmes de « soutien à la relève » sans création proportionnée de postes stables (Fassa et Kradolfer 2013) et la dynamique « d’européanisation » de la recherche (Felli et al. 2007) au cours des années 1990‒2000, ainsi que la création d’une multitude de nouveaux statuts et postes non-stables au sein des universités suisses ont largement contribué à remettre en question cet équilibre fragile et à créer une bulle postdoctorale de grande ampleur à l’échelle nationale (Leemann et al. 2010). Comme le montre la Figure 1, depuis 1980, le ratio entre les personnels académiques non-stables assistant·e·s et « collaborateur·trice·s scientifiques » (soit l’ensemble des assistant·e·s diplômé·e·s, des premier·e·s assistant·e·s, maître·sse·s assistant·e·s et postdoctorant·e·s embauché·e·s sur des projets de recherche ponctuels) et les personnels académiques stables et/ou occupant une poste stabilisable (professeur·e·s titulaires de chaires et « autres enseignant·e·s » ici7) a ainsi été divisé par 2. Ces métamorphoses n’ont pas été sans conséquences quant aux représen- tations et pratiques des jeunes chercheur·e·s évoluant dans l’espace acadé- 6 Dans la tradition allemande, les Privatdozent sont des enseignant·e·s doté·e·s d’une « habilitation » (donnant théoriquement le droit de candidater sur des postes de professeur·e·s), pouvant proposer des cours dans une université sans être rémunér·e·s par l’État néanmoins. Ce statut particulièrement précaire, est généralement considéré comme la « position d’attente » classique pour un poste professoral, bien que les chances objectives de nomination sur une chaire aient toujours été très réduites (Schultheis 2000, 59). 7 On regroupe derrière cette étiquette les maître·sse·s d’enseignement et de recherche (MER) et les professeur·e·s assistant·e·s notamment. 254 mique helvétique. En effet, face à cette réduction importante des chances de « survie » pour accéder à un poste académique stable, un nombre croissant de postdoctorant·e·s – et en particulier ceux et celles formé·e·s en Suisse – se détournent de la voie académique pour viser des postes dans les entreprises ou l’administration, plus rémunérateurs et garants d’une progression de carrière plus linéaire que les postes universitaires (Bataille et al. 2017). Néanmoins, si le contexte helvétique apparaît avoir été particulièrement en pointe quant aux différentes métamorphoses qui ont touché la plupart des espaces académiques nationaux depuis la fin des années 1990, les incidences de ces changements sur l’organisation traditionnelle des carrières et des recrutements à l’univer- sité restent relativement mal connues. Le cas suisse apparaît ainsi comme un exemple particulièrement intéressant pour observer les possibles hybridations entre modèles de carrières, au cœur de notre problématique. 11.3.2 Une analyse tridimensionnelle des parcours post-doctoraux Pour analyser les parcours de nos enquêté·e·s et croiser données qualitatives et quantitatives, nous proposons d’utiliser une analyse tri-dimensionnelle. Premièrement, grâce à une analyse de correspondances multiples spécifique (ACM) réalisée à partir des variables relatives à la carrière professionnelle et aux conditions de travail, nous traçons les contours d’un « espace des parcours postdoctoraux » en Suisse et dressons une typologie des différents possibles professionnels qui s’ouvrent aux jeunes docteur·e·s une fois décroché leur premier postdoc en Suisse. Deuxièmement, il s’agit de repérer, grâce à la projection de modalités supplémentaires dans le plan factoriel, les propriétés sociales individuelles qui semblent prédisposer les individus à s’engager sur un type de parcours postdoctoral spécifique. Enfin, dans un troisième temps, nous opérons une « mise en espace social » (Darmon 2008, 152) des expériences de mobilité des personnes que nous avons interviewées, en proposant une analyse de leur discours par le prisme du poste qu’elles occupent dans l’espace constitué grâce à l’ACM – en utilisant la technique présentée par Renisio et Sinthon (2014) de projection des interviewé·e·s dans notre plan factoriel en tant qu’individus supplémentaires ne participant pas à l’orientation des axes. Les modalités actives utilisées dans notre ACM sont présentées dans le Tableau 1. Elles renvoient tout d’abord à des variables permettant de retra- cer la carrière de nos interviewé·e·s, telles que leur âge académique, le lieu de formation doctorale ou encore le type de poste occupé au moment de l’enquête. Elles permettent également de capturer d’autres indicateurs a priori liés à l’avancement de la carrière et à l’accès à un poste académique stable, comme l’accumulation de capital scientifique ou l’étendue du capital social 255 au sein de l’espace académique. Les modalités trop peu fréquentes pour être utilisées dans l’analyse sont indiquées en italique. Nous avons également écarté les non-réponses des modalités actives – elles figurent donc également en italique ci-dessous. Ces premiers chiffres appellent quelques commentaires préliminaires. Notamment, les informations quant au lieu de soutenance de la thèse per- mettent de se faire une idée des flux qui alimentent potentiellement la bulle postdoctorale suisse : on voit ici qu’il s’agit majoritairement de migrations Nord-Nord, d’un pays de l’UE ou d’un pays d’Amérique du Nord vers la Tableau 1: Modalités actives de l’ACM Variables Modalités % Nombre d’années de postdoc 0–3 21,7 4–6 45,7 7–9 21,7 + 9 10,1 NR 0,7 Pays/Université de Soutenance Suisse 32,6 UE 50,7 Am-Nord 10,1 Autre 5,8 NR 0,7 Situation professionnelle Hors académie 7,2 Position académique permanente 15,2 Position académique temporaire 74,6 Sans emploi 2,2 NR 0,7 Conflit directeur de these Oui 13,0 Non 87,0 Articles peer-reviewed 0–1 5,1 2–5 27,5 5–10 28,3 + 10 30,4 NR 8,7 Membres de réseaux académiques Oui 76,8 Non 23,2 NR 0 N 138 Source : Projet GARCIA, enquête en ligne. Lecture : Dans l’échantillon de répondant·e·s à notre enquête par internet, 74,6 % des personnes ques- tionnées occupaient au moment de l’enquête un poste académique non-stable. 256 Suisse. Les docteur·e·s des universités suisses sont également bien représenté·e·s dans les rangs de nos enquêté·e·s (32,6 % en tout). Les modalités illustratives réunies dans le Tableau 2 renseignent quant à elles sur des dimensions du parcours social antérieures ou parallèles à la carrière proprement professionnelle (le sexe ou encore le diplôme du/de la partenaire) ou encore sur la discipline de spécialisation. Tableau 2 : Modalités illustratives de l’ACM Variables Modalités % Sexe Femmes 60,6 Hommes 39,4 Discipline Sciences Humaines et Sociales (SHS) 26,8 Sciences, Technologies et Médecine (STM) 73,2 Diplôme partenaire Doctorat 29,0 Enseignement supérieur 34,8 Primaire ou secondaire 7,2 Célibataire 29,0 N 138 Source : Projet GARCIA, enquête en ligne. Lecture : Dans l’échantillon de répondant·e·s à notre enquête par internet, 29 % des personnes sont en couple avec un individu titulaire d’un doctorat. Il apparaît très nettement que les sciences dites « dures » (STM) sont plus touchées par le phénomène de la multiplication des contrats académiques courts, comme le montre la sur-représentation des enquêté·e·s ayant travaillé dans le département STM de l’université étudiée. 11.4 L’espace des parcours postdoctoraux en Suisse La Figure 2 représente les modalités participant plus que la moyenne à l’orien- tation des deux premières dimensions de l’ACM, qui expriment respective- ment 11,8 pour cent et 9,4 pour cent de la variance du nuage de points. La première dimension oppose les nouveaux entrant·e·s, engagé·e·s depuis trois ans ou moins sur la voie postdoctorale, ayant à ce stade peu publié d’articles dans des revues à comité de lecture, souvent embauché·e·s sur des postes académiques temporaires, et les postdoctorant·e·s plus confirmé·e·s, ayant soutenu depuis plus de neuf ans leur thèse et ayant par ailleurs accumulé un important capital scientifique. 257 Figure 2: L’espace des profils postdoctoraux en Suisse (dimensions 1 et 2) Dir_conf 1.5 Aut Hors_acad 1.0 NetMemb 0.5 7−9 2−5 + 10 0.0 Acad_temp Dir_nconf Acad_perm 0−3 −0.5 UE −1.0 0−1 + de 9 PhD_Aut NR −1.5 −1.0 −0.5 0.0 0.5 1.0 Dim1 (11.8%) Variables Membre réseaux Nombre articles académiques peer-reviewed Pays soutenance PhD Relations directeur·rice·s de thèse Nombre années postdoc Pays de travail Position professionnelle Source: Projet GARCIA – enquête en ligne. Lecture : L’analyse de correspondances multiples permet de dégager deux grandes oppositions structu- rantes dans les profils post-doctoraux – entre ceux et celles qui ont des carrières longues, de nom- breuses publications scientifiques et qui accèdent à un poste académique permanent (Hors Suisse, dans l’Union européenne bien souvent) et les autres d’une part ; entre ceux et celles qui ont eu des conflits avec leur directeur de thèse et qui quittent le milieu académique et les autres d’autre part. Dim2 (9.4%) 258 Une partie des postdocs « avancé·e·s » a décroché un poste stable dans un éta- blissement suisse d’enseignement supérieur, mais une majorité des chercheur·e·s ayant accédé à un poste stable s’est au contraire établie en dehors des frontières suisses, dans un pays de l’Union Européenne. Il apparaît ainsi que le passage au niveau postdoctoral dans une université suisse constitue un tremplin pour briguer par la suite des postes stables dans d’autres pays. Ce type de stratégie est assez courant chez les postdoctorant·e·s issu·e·s des pays frontaliers de la Suisse (l’Italie, mais surtout la France ou l’Allemagne), ainsi que nous avons pu le montrer dans une analyse plus systématique (Bataille 2016). Pour ces chercheur·e·s, un passage de quelques années par une université suisse constitue une solution de choix pour acquérir une expérience à l’étranger à la fois confortable sur le plan économique, moins difficile à mettre en place sur le plan logistique qu’une migration en Amérique du Nord, et intéressante sur le plan scientifique – dans la mesure où la condensation d’établissements d’enseignement supérieur de très haut niveau sur un territoire restreint est propre à susciter une relative émulation. En outre, le caractère extrêmement sélectif du recrutement professoral suisse dissuade souvent les nouveaux et nouvelles arrivé·e·s de considérer leur lieu de postdoc comme un espace de recrutement à long terme. Une opposition importante apparaît le long de l’axe figurant la deuxième dimension de l’ACM, entre ce premier profil de parcours menant à la stabi- lisation et des carrières également longues mais débouchant sur des postes non-académiques, en Suisse ou à l’étranger. Cette dernière opposition apparaît largement liée à l’intégration dans le milieu académique et au fait de bénéficier, ou non, d’appuis importants au sein de l’institution. Les personnes rapportant un conflit avec leur directeur ou directrice de thèse sont par exemple davan- tage représentées au sommet de l’axe 2. Si le premier axe apportait certains arguments pour remettre en question le modèle traditionnel de carrière dans l’espace académique helvétique – où le soutien durable du directeur ou de la directrice de thèse est une condition sine qua non pour espérer être nommé·e sur un poste de professeur·e (Rolle 2016) – ce deuxième axe semble plutôt attester d’une relative longévité de ce modèle du survivor. Le pays de soutenance de la thèse apparaît également comme un élément important. Parmi les docteur·e·s de nationalité américaine ou canadienne, peu sont ceux et celles qui ont opté pour un poste hors académie. Cela s’ex- plique en partie par les politiques migratoires en Suisse, qui rendent difficile l’établissement à long terme des ressortissant·e·s non-européen·ne·s. De fait, comme nous le verrons dans les cas analysés par la suite, ces doctorant·e·s non-européen·ne·s qui ne disposent pas toujours de solution de retour dans 259 leur pays d’origine, sont souvent amené·e·s à enchaîner les contrats de courte durée dans des universités suisses ou européennes. Analysons maintenant comment se répartissent les individus au sein de cet espace des parcours postdoctoraux. Afin de rendre compte plus finement de la dispersion des profils sociologiques, nous avons choisi de repérer, à l’échelle du nuage d’individus constitué via les résultats de l’ACM, les zones où se concentre, pour chaque modalité, la majeure partie des enquêté·e·s. Ces zones sont « résumées » (Le Roux et Rouanet 2010) par des ellipses médianes délimitant une surface circulaire dans laquelle se trouve au moins 50 pour cent des personnes concernées. L’axe le plus long de l’ellipse correspond à l’orientation générale du nuage d’individus. L’axe le plus court est propor- tionnel à la variance du nuage. La Figure 3 représente la localisation des répondant·e·s au sein de notre espace des parcours, et ce en fonction de leur champ disciplinaire de spécialité. On voit ici qu’il y a très peu de différences si l’on ne considère que la variable en question : les deux ellipses ont une forme et un poste similaires dans le plan factoriel, témoignant d’un alignement des logiques structurant les carrières dans les sciences humaines et sociales sur celles à l’œuvre dans les sciences dites « dures » dans l’espace helvétique. Le sexe des répondant·e·s induit peu de différences quant à la localisation sur le plan factoriel. C’est surtout dès que l’on croise la variable sexe et le niveau de diplôme du/de la conjoint·e (Figure 4), que l’on voit apparaître d’importantes différences entre les répondantes et les répondants. Être en couple avec une personne disposant d’un doctorat apparaît plutôt comme un frein à la stabilisation, tant la majeure partie des individus dans cette situation se situe à l’est du graphique, quel que soit le sexe. Les condi- tions conjugales favorisant la stabilisation distinguent en revanche nettement femmes et hommes : les postdocs hommes ayant accédé à un poste stable sont plus souvent en couple avec une compagne diplômée du secondaire ou du primaire. Les femmes les plus proches du pôle regroupant les enquêté·e·s ayant décroché un poste stable sont en revanche plus souvent célibataires que les hommes accédant à ce même type de poste. S’il n’est pas directement question de mobilité ici, on peut voir dans cette différenciation sexuée des configurations conjugales semblant faciliter l’accès à la stabilisation un effet indirect de l’injonction à partir pendant une plus ou moins courte période à l’étranger en séjour postdoctoral directement après la thèse, élément souvent posé comme une condition pour accéder à un poste académique stable dans le contexte suisse. Être mobile pour partir avec son conjoint est d’autant moins aisé dans les couples à double carrière, une configuration plus fréquente chez les femmes fortement diplômé·e·s. Aussi, on peut penser que celles qui 260 Figure 3 : Nuages des individus – disciplines (dimensions 1 et 2) 1.5 1.0 0.5 0.0 −0.5 −1.0 Discipline −1.5 STEM SHS −2.0 −1.5 −1.0 −0.5 0.0 0.5 1.0 Dim1 (11.8%) Source : Projet GARCIA – enquête en ligne. Lecture : Les deux sous-nuages de points en fonction de la spécialité disciplinaire (STEM ou SHS) occupent une place similaire dans le plan factoriel. entament néanmoins une mobilité postdoctorale sont susceptibles d’expéri- menter de fortes tensions conjugales pouvant aboutir à une rupture, comme c’est le cas chez nombre de nos enquêtées. Pour les hommes, plus souvent en situation de « male breadwinner » – ce modèle restant largement dominant Dim2 (9.4%) 261 Figure 4: Nuage des individus – sexe et niveau de diplôme du/de la par- tenaire (dimension 1 et 2) Femmes Hommes 1.5 1.0 0.5 0.0 −0.5 −1.0 −1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0 Dim1 (11.8%) Dip. part. Célibataires Études Enseignement PhD secondaires supérieur Source : Projet GARCIA – enquête en ligne. Lecture : En fonction du sexe des répondant·e·s, les sous-nuages de points en fonction du niveau de diplôme du partenaire sont disposés différemment sur le plan factoriel. Chez les hommes notam- ment, le fait d’être en couple avec un·e partenaire un peu moins diplômé semble être plus souvent associé à l’accès à un poste académique stable. dans le contexte suisse (Le Feuvre 2015) –, la situation est différente dès lors que leur potentielle partenaire peut mettre entre parenthèses son activité professionnelle pour les accompagner pour un séjour à l’étranger. 11.5 Entre logiques locales et mobilités internationales : des expériences hybrides Une fois dégagées les grandes lignes directrices qui structurent les carrières de nos enquêté·e·s, nous pouvons à présent analyser plus finement les implications subjectives de la mobilité postdoctorale – et de ses importantes variations de signification au regard de la situation occupée par les personnes rencontrées. La mobilité académique étant au cœur de notre texte, nous nous focalisons dans la suite de l’énoncé sur les problématiques relatives à cette question – et Dim2 (9.4%) 262 Figure 5 : Localisation des interviewé·e·s dans le plan 1.5 1.0 0.5 Stephany Alberto Georges 0.0 Barbara −0.5 −1.0 −1.5 Interviews Non Oui −2.0 −1.5 −1.0 −0.5 0.0 0.5 1.0 Dim1 (11.8%) Source : Projet GARCIA – enquête en ligne. Lecture : Dans le plan factoriel obtenu grâce aux deux premières dimensions de l’ACM, nos interviwé·e·s occupent des places différentes. Barbara et Georges sont plutôt proche du pôle « stabilisé » sur un emploi académique. Stephany et Alberto sont quant à eux plus proche du pôle « non-stable ». Dim2 (9.4%) 263 laissons en suspens l’analyse des trajectoires de ceux et celles qui quittent la carrière académique8. La Figure 5 présente la localisation de nos interviewé·e·s dans le plan factoriel dessiné à partir des deux premières dimensions de notre analyse. Les quatre prénoms (des pseudonymes) correspondent aux 4 cas que nous présentons ci-après. Ces cas ont été choisis au regard de la diversité des postes qu’ils permettent d’explorer, mais aussi en fonction de l’éclairage qu’ils apportent quant à l’intrication des questions liées à la conjugalité et à la mobilité, abordées dans les analyses qui précèdent. 11.5.1 Des États-Unis à la Suisse, le parcours de Stephany entre deux continents Stephany est d’origine nord-américaine. Après un doctorat en sciences natu- relles obtenu dans une prestigieuse université privée, elle quitte les États-Unis, pour prendre un poste de première assistante dans une université de Suisse romande. Le lieu de cette mobilité en Europe résulte d’un choix de couple, puisque son mari, postdoctorant de nationalité belge, souhaite travailler dans un environnement francophone. Les conjoints réalisent ainsi une recherche d’emploi simultanée, postulent à différentes annonces, et arrêtent finalement leur choix sur la Suisse en croisant leurs intérêts de recherche respectifs et la volonté de se trouver dans la même zone géographique. Les deux emplois se trouvent à une heure de trajet, et le couple loue ensemble une habitation à mi-distance. La mobilité de Stephany est également une mobilité disciplinaire ; en candidatant en Suisse, la jeune chercheuse postule en effet dans une disci- pline différente – bien que voisine – de sa spécialisation d’origine. Stephany contextualise cette décision en évoquant notamment les discriminations de genre et l’hostilité auxquelles elle fait face dans son laboratoire aux États- Unis, dans un milieu de sciences dures qu’elle décrit comme particulièrement hostile aux femmes. « Fatiguée de se battre », Stephany prend alors la décision de réaliser son postdoctorat en sciences de l’environnement, un milieu dis- ciplinaire qu’elle identifie comme moins discriminant envers les femmes, et plus féminisé. Sa prise de poste en Suisse ne se fait cependant pas sans heurts. Stephany est déconcertée par le caractère flou de son contrat, l’absence de cahier des charges, et ne comprend que tardivement – à défaut de maîtriser les 8 Pour une analyse des cas des postdoctorant·e·s suisses qui se reconvertissent hors académie, nous renvoyons à notre article traitant plus spécifiquement ce sujet (Bataille et al. 2017). 264 codes du contrat et de la langue française – être embauchée à temps partiel, ou encore être responsable de tâches d’enseignement. Très impliquée dans ses recherches et la conduite de nombreuses activi- tés connexes à l’interface entre science, culture et société, Stephany dresse cependant un portrait amer du système universitaire, qui nourrit, par le manque de postes et la compétition qui en résulte, la précarité des jeunes chercheur·e·s. Pour Stephany, le monde académique repose sur un système dysfonctionnel, dans lequel les postdocs constituent une « main d’œuvre bon marché », qui donne beaucoup d’elle-même sans pouvoir espérer en retour une sécurité professionnelle. Les chercheur·e·s s’agglomèrent ainsi dans cet espace « obstrué », dans lequel les individus ayant enchaîné dix années de postdocs rivalisent non seulement entre eux et elles, mais également avec de nouvelles générations d’entrant·e·s. En outre, Stephany souligne que l’accès à un poste professoral ne vient pas toujours signer la fin de cette incertitude. En tant qu’américaine, sa situation d’emploi est d’autant plus précaire, que son permis de travail prendra fin, simultanément à son contrat, au bout de cinq ans. Cette configuration, commune pour ce type de poste, la conduit à mener de front d’intenses activités académiques propres à la fin d’une recherche – clore l’expérimentation et publier des articles – et la recherche d’un nouvel emploi qui commencerait immédiatement. La jeune chercheuse travaille au laboratoire en journée mais consacre également l’essentiel de ses soirées, et la moitié de ses week-ends, à la réalisation d’un projet parallèle, financé par le Fonds national suisse de la recherche scientifique (FNS). Dans un contexte marqué par un emploi du temps chargé, et l’absence de sécurité professionnelle, le projet de parentalité du couple, qui souhaiterait avoir un enfant, ne lui paraît pas réalisable. Stephany considère que ses principales opportunités d’emploi académique se situent aux États-Unis. Cette destination n’est cependant envisagée par son conjoint que dans le cas où Stephany décrocherait « le job idéal », une condition qu’elle juge très limitante. Bien que passionnée par son activité quotidienne de chercheuse, elle envisage de quitter le milieu académique sans trouver pourtant de débouchés aisés dans son domaine de prédilection. Pour cette postdoctorante, le système actuel nuit aux individus mais aussi à la science elle-même : la précarité des jeunes chercheur·e·s favorise des attitudes de rivalité au sein d’une même équipe aux dépens de la collaboration, et conduit nombre de ses collègues à quitter le milieu académique pour travailler sur des projets peu utiles à la société, mais offrant de meilleures conditions salariales, de sécurité, et de reconnaissance. Les caractéristiques de Stephany la positionnent dans le cadran nord-est de notre plan factoriel (Figure 5). Son récit permet de mieux appréhender 265 quelques-unes des logiques qui président au cheminement de cette jeune chercheuse en sciences dures de part et d’autre de l’Atlantique. Évoluant sur un poste instable près de quatre années après la fin de sa thèse, la jeune cher- cheuse se perçoit comme une travailleuse invisible au devenir précaire plutôt que comme une chercheuse autonome, bâtissant progressivement sa carrière académique en relation avec des acteurs et actrices locaux et internationaux. La dépendance financière à ses superviseurs, la difficulté à faire valoir ses propres résultats, le manque d’allié·e·s et de soutiens locaux tout comme les limitations qui encadrent son permis de travail la positionnent davantage comme une source de main-d’œuvre provisoire et de passage, que comme une chercheuse indépendante, intégrée sur le marché local, et dont l’interna- tionalité du parcours constituerait un avantage en soit. Ce parcours féminin illustre enfin de quelle manière le fait d’être en couple avec un conjoint lui aussi doté d’un diplôme de doctorat et évoluant dans un marché du travail académique similairement restreint, est susceptible de contraindre mobilité et opportunités d’emploi. 11.5.2 Bouger pour se faire un réseau, le parcours de Barbara aux frontières de la France Après un doctorat en sciences sociales réalisé dans le sud de la France, Bar- bara déménage à Bordeaux – où elle effectue un premier postdoctorat d’un an – puis s’expatrie dans une université de Suisse romande pour un second postdoctorat de deux ans. Pendant cette période, elle maintient des liens de collaboration étroits avec ses ancien·ne·s collègues et superviseur·e·s. À 35 ans, elle est classée première à plusieurs auditions sur des postes français, et retourne en France en tant que maîtresse de conférences. Barbara explique que sa décision de quitter sa région à la fin de sa thèse repose sur le manque de perspectives académiques locales, et plus particuliè- rement sur des règles tacites l’excluant d’un recrutement dans son université d’origine. Barbara est en couple avec un chercheur français, évoluant dans une discipline rattachée aux STM. Devant la difficulté à trouver des postes doctoraux puis postdoctoraux à proximité immédiate l’un de l’autre, les conjoint·e·s vivent pendant de nombreuses années à distance, tout en privilé- giant des emplois situés dans un espace géographique circonscrit et concerté : les laboratoires de recherche les plus attractifs dans leur domaine respectif, mais situés sur d’autres continents sont ainsi écartés au profit des postes situés en France métropolitaine et dans les pays européens, une configuration qui permet au couple de se voir « régulièrement », à hauteur de deux fois par mois. Aux yeux de la jeune chercheuse, le manque de postes, et les ajustements que 266 cette limitation génère, ont été à l’origine de sacrifices dans sa vie privée : celui de vivre à distance de son partenaire, mais aussi celui de repousser le projet de devenir parents. Barbara raconte cependant mettre en partie à profit cette vie de couple à distance. Habiter loin de son conjoint lui permet de travailler fréquemment de longues heures, de 9h à 21h à l’université, et de consacrer de nombreux week-ends à l’avancement de sa recherche. Cette configuration lui a permis de « jouer le jeu du postdoctorat : c’est-à-dire d’optimiser le temps (pendant un ou deux ans) », un choix qui lui paraît valoir la peine, dans la mesure où le contexte académique français permet une stabilisation relativement rapide. Barbara estime que sa sélection à plusieurs auditions – et, in fine, sa stabilisation – sont notamment liées à des phénomènes d’interconnaissance, facilités par le fait que sa discipline est un champ étroit, un « petit monde » dans lequel elle s’est fait connaître en multipliant les villes et laboratoires de travail, en créant des liens avec des individus différents et plus généralement en privilégiant des activités de mise en réseau, de conférence, et de collabo- rations multiples. Sélectionnée sur plusieurs postes, elle prend en compte deux critères prin- cipaux dans son choix final : le laboratoire d’accueil, mais aussi la possibilité de se rapprocher géographiquement de son conjoint, stabilisé sur un poste similaire. Le couple habite désormais à trois heures de distance et projette de poursuivre ce rapprochement en demandant régulièrement des mutations. Les caractéristiques de Barbara la positionnent à l’ouest de l’espace des cheminements post-doctoraux (Figure 5). Après deux postdocs en France et en Suisse, Barbara occupe désormais un poste stabilisé dans son pays d’origine. Pour Barbara, la période postdoctorale en Suisse est synonyme d’un très fort investissement professionnel, qui se fait aux dépens de la sphère privée. Si elle parle de « sacrifice », cette période est également associée à un vécu subjectif positif – notamment celui du « jeu » – qui apparaît lié aux caractéristiques du système de carrière dans lequel elle se projette : le fait de travailler en soi- rée et en week-ends, et à distance du conjoint, lui apparaissent comme des concessions acceptables dans la mesure où elles s’effectuent sur une période limitée « d’un ou deux ans », dans un système français qui laisse espérer (et ici, a permis) une stabilisation rapide. Barbara conçoit indépendamment ses deux postdocs en France et à l’étranger comme des moyens d’augmenter son capital scientifique et social, de « construire son réseau ». Si la chercheuse met à distance une représentation linéaire et locale de la carrière qui consisterait à être stabilisée à l’issue du doctorat dans l’université d’origine, elle inscrit sa trajectoire dans une logique relativement semblable : le postdoctorat est vécu comme une étape courte 267 et transitoire et la stabilisation intervient au sein d’un tissu professionnel où elle est déjà largement identifiée, et dont elle est partie prenante. La logique du « recrutement local » se déploie ici non plus à l’échelle institutionnelle, mais davantage à une l’échelle nationale, un changement de magnitude qui apparaît facilité par le fait que sa discipline constitue « un petit monde », un champ étroit. 11.5.3 Du chercheur étranger au chercheur local ; la trajectoire d’Alberto au prisme des normes de carrière helvétiques Alberto quitte l’Italie à la suite de son master avec pour objectif de trouver un financement de doctorat viable dans un pays voisin. Il s’installe alors en Suisse romande où il rencontre sa compagne ; le couple a un enfant. À la suite du doctorat, Alberto souhaite faire un séjour à l’étranger afin de de se conformer aux normes de carrière suisse, mais renonce suite à l’ultimatum de sa partenaire : « C’était si tu pars, on se quitte » La compagne d’Alberto, elle-même jeune docteure, rejette l’idée de quitter la Suisse et quitte le milieu académique pour occuper un poste fixe en Suisse romande. Sur les conseils de son directeur de thèse, Alberto adopte plusieurs stra- tégies pour compenser l’absence de mobilité internationale dans son dossier académique. D’une part, il change sciemment d’axe de recherche au sein de sa discipline afin de témoigner de sa flexibilité sur le plan intellectuel. D’autre part, à l’issue de son postdoctorat, il postule pour une bourse Ambizione – une des rares bourses du FNS pour laquelle un antécédent de mobilité est « souhaité mais pas obligatoire » – et ce dans une ville de Suisse alémanique, germanophone, qui selon son superviseur, garantit « tout le choc de l’étranger ». Si Alberto considère qu’une mobilité à l’étranger lui aurait permis d’élar- gir son réseau, de parfaire son anglais et de faire connaissance avec d’autres manières de travailler, il estime également que le fait d’être resté en Suisse n’a en aucun cas limité son développement scientifique. En outre, Alberto rejette le conseil formulé par ses pairs et sa hiérarchie d’effectuer une mobilité factice dans un pays proche, une pratique selon lui courante. Alberto décrit son dossier académique comme très compétitif, il publie dans Nature et dans plusieurs journaux majeurs de son champ de recherche. Sa relative séden- tarité postdoctorale lui est cependant systématiquement reprochée dans ses dossiers de candidature et limite ses opportunités de carrière. Pour Alberto, la réussite universitaire se détermine en grande partie en fonction de la marge de manœuvre dont les chercheur·e·s disposent sur le plan privé. En outre, il regrette que l’absence d’un unique critère dans son CV, l’exclue d’un très large éventail de financements ou de postes. Il est ainsi très critique vis-à-vis non 268 seulement de la norme de mobilité, mais aussi de l’uniformité d’un système d’évaluation qui participe selon lui à standardiser les parcours scientifiques. À 38  ans, Alberto se montre désormais clairement pessimiste sur ses chances de stabilisation ; il se projette à contrecœur dans un rôle managérial à l’extérieur du monde universitaire et montre une certaine amertume envers ses employeurs successifs, des universités qui bénéficient de son travail, de l’obtention de ses bourses de recherche, ou encore de son exposition média- tique au moment de publier dans Nature, sans pour autant lui offrir une sécurité professionnelle. Les caractéristiques d’Alberto le positionnent à l’est de notre espace des cheminements postdoctoraux (Figure 5). Tout comme Stephany, Alberto est un chercheur étranger qui évolue dans une discipline de sciences dures. Cependant, ce jeune chercheur, au bénéfice d’un doctorat suisse, livre le récit d’un chercheur local et inséré. En particulier, Alberto connaît bien les normes de carrière suisses, les pôles de recherche attractifs dans son domaine et les possibilités de financement. Il bénéficie en outre du soutien et des conseils de ses pairs, de son superviseur de postdoctorat et de son ancien directeur de thèse, qui l’aident à se positionner sur le marché local, à construire son CV à moyen et à long terme, et à choisir un laboratoire stimulant pour la réalisation de son postdoctorat. Les difficultés liées à la double carrière, qui s’expriment ici en défaveur d’un profil masculin, renvoient en partie à la même logique que celle déjà décrite dans le cas de Stephany et de Barbara : dans les trois cas, le départ est envisagé à l’étranger, mais ne bénéficie pas de l’adhésion automatique du ou de la conjointe, également très qualifié·e, et qui compose notamment avec son propre agenda professionnel. Dans les études de cas que nous avons proposées, le haut niveau d’étude du conjoint de la conjointe peut être analysé comme un frein à la carrière (Figure 5) à partir de différents mécanismes, souvent cumulés : dans les trois premiers cas, le ou la partenaire – sensiblement du même âge professionnel – est en effet en activité, sur un poste également qualifié et rémunérateur. Dans le cas de Stephany et de Barbara, on sait que le partenaire doit lui-même composer avec la difficulté à construire sa propre trajectoire académique et choisir en toute liberté son lieu de travail. En outre, le projet de mobilité professionnelle de Stephany, Alberto, ou Barbara, offre peu de garanties : le poste envisagé par ces chercheur·e·s est précaire, temporaire, et ne garantit pas l’accès futur à une sécurité professionnelle dont pourrait partiellement bénéficier le ou la partenaire. 269 11.5.4 Entre local et international : le parcours au masculin de Georges Georges représente un des cas les plus emblématiques de l’hybridation des deux modèles de carrières, mêlant mobilité internationale et forts ancrages locaux. Il permet d’étayer l’idée que c’est l’accumulation des atouts relatifs à chacun de ces deux modèles qui permet d’accéder relativement rapidement à un poste stable dans le contexte étudié. Son cas permet également de montrer comment le fait d’occuper une place de « pourvoyeur principal de ressources » au sein de ses couples successifs est également un atout non négligeable pour emprunter ce type de parcours. Georges réalise son doctorat en sciences sociales dans une université de Suisse romande. La dernière année est réalisée au Québec, grâce à une bourse mobilité du FNS. À son retour, Georges obtient rapidement – grâce son directeur de thèse – un poste stable de recherche à 60 pour cent dans une haute école suisse, puis est contacté par un professeur d’université voisine, avec lequel il a déjà eu l’occasion de collaborer pendant sa thèse, et qui lui propose un poste de maître-assistant sur le temps de travail restant. S’il estime que le poste de recherche en haute école ne bénéficie pas du même prestige social que l’université, ce poste lui permet de jouir d’une sécurité financière durable, de trouver un bon équilibre avec le poste – instable – de maître-assistant, et lui retire durablement la pression d’avoir à trouver un poste permanent. Cette période principalement consacrée à la recherche lui permet d’enrichir son CV, et de publier de manière satisfaisante. Pourtant, Georges estime que son dossier est loin d’être assez compétitif pour espérer obtenir un des rares postes professoraux dans une université suisse. Quatre ans après sa thèse, il apprend la mise au concours d’un poste en tenure track dans l’université québécoise avec laquelle il a gardé des contacts, et postule sur un coup de tête, avec le soutien de celui qui encadre son postdoctorat. Le jeune professeur repart ainsi au Canada, s’intègre parfaitement à son environnement et est promu sur un poste stabilisé à l’issue d’un contrat de cinq ans, mais saisit cependant un an plus tard l’opportunité de rentrer dans son pays natal, à la faveur de la mise au concours d’un poste de professeur associé. Soutenu notamment par son ancien superviseur, avec lequel il a maintenu des liens de collaboration scientifique étroits et qui siège dans le comité de recrutement, Georges réintègre ainsi l’institut universitaire dans lequel il avait fait, dix ans plus tôt, ses premiers pas de postdoctorant. Georges estime que ce sont particulièrement de « bons contacts » et les conseils stratégiques dont il a bénéficié grâce à son mentor en Suisse – que ce soit sur les activités à privilégier, le réseau à construire, les revues dans les- 270 quelles publier – qui lui ont permis de « tirer son épingle du jeu ». Les postes postdoctoraux constituent à ses yeux la marque d’un système malade, qui « mène les gens à se déchirer entre eux » sans offrir une chance significative de stabilisation ; les rares chercheur·e·s qui parviennent à survivre dans cet espace précaire ne sont alors pas forcément les « bon·ne·s » chercheur·e·s mais ceux qui connaissent bien le système dans lequel ils évoluent et « adhèrent mieux aux règles », ceux et celles qui sont de « bons stratèges ». Georges décrit enfin comment la progression de sa carrière est facilitée par la flexibilité de ses partenaires successives : la première accepte de mettre entre parenthèses ses propres projets professionnels pour le suivre au Canada, une situation qui génère cependant à moyen terme la rupture du couple ; la seconde, docteure non stabilisée, suit Georges en Suisse avec le projet de trouver sur place des opportunités intéressantes au niveau postdoctoral. Les caractéristiques de Georges le positionnent très clairement à ouest du plan factoriel (Figure 5). Il s’agit d’un parcours en sciences sociales, qui, si l’on s’arrête à la Figure 2, renvoie à un mécanisme de stabilisation locale différé : Georges est natif et diplômé de Suisse, mais surtout le poste permanent qu’il vient d’obtenir au moment de l’entretien se trouve dans le même institut que celui dans lequel il a effectué ses premiers mandats en tant que postdoc. Dix ans séparent ces deux postes, selon une temporalité qui rappelle une des caractéristiques du modèle de carrière du survivor, pourtant, la trajectoire de Georges s’éloigne par plusieurs aspects des modes traditionnels de recrutement. Le poste de recherche à l’extérieur de l’université, le renoncement précoce à postuler sur un poste professoral local et le choix de s’insérer durablement dans un marché académique canadien perçu comme moins sélectif, sont vécus par Georges comme des configurations alternatives relativement éloignées des standards de réussite académique helvétique. Ces configurations peuvent être analysées comme résultant d’une hybridation entre les tendances globales de dé-standardisation et d’internationalisation des carrières, et les modèles de recrutement nationaux. Après la thèse, Georges enchaîne en effet plusieurs mandats postdoctoraux temporaires et à temps partiels à l’université, tout en travaillant sur un poste de recherche extérieur au monde universitaire, moins prestigieux mais stable – et qui lui donne paradoxalement certains avantages vis-à-vis de collègues postdoctorant·e·s plus précaires. Au Canada comme en Suisse, le recrutement de Georges est facilité par la présence de soutiens locaux, avec lesquels il a déjà eu l’occasion de collaborer au gré de ses mobilités et mandats successifs. Enfin, très pessimiste sur ses chances d’être stabilisé en Suisse en l’absence d’offres de poste suffisantes, Georges se projette rapidement dans un marché du travail transnational : la mobilité vers l’étranger n’est pas perçue, comme dans le cas d’Alberto, comme une case à cocher sur le CV 271 avant de postuler en Suisse, ou, comme dans le cas de Barbara, comme une étape transitoire préludant à une stabilisation rapide dans le pays de naissance, mais comme une modalité durable d’emploi, induisant un changement radical de lieu de vie pour le chercheur comme pour sa partenaire. L’étude qualitative des entretiens permet ainsi de pointer l’hybridation des différents modèles. Par exemple, la trajectoire de Georges, caractérisée par une stabilisation rapide, s’affranchit partiellement de ce modèle en prenant appui sur les possibilités offertes par l’internationalisation du marché de l’emploi académique. Le cas de Barbara vient quant à lui souligner, par comparaison avec celle de Stephany et d’Alberto, comment les modèles et dispositifs de carrière locaux sont reconfigurés ou détournés au contact de nouvelles popu- lations d’entrants : Barbara perçoit son expérience postdoctorale en Suisse, comme une étape courte, peu connectée à l’institution et au marché local, mais au contraire répondant aux attentes du marché académique français dont elle est issue. Il s’agit enfin de souligner à quel point les phénomènes de dé-standar- disation et d’internationalisation des trajectoires académiques participent à reconfigurer la gestion des doubles carrières, et à bousculer les frontières entre sphère professionnelle, privée et familiale. Barbara, Stephany, et Alberto, montrent tou·te·s les trois des stratégies très différentes de gestion parallèle de la conjugalité et de la carrière en vivant à distance, en partant ensemble pour une destination attractive commune ou en donnant priorité à l’ancrage géographique de l’un des deux conjoints. Dans ces trois cas, le suivi du ou de la conjoint·e est d’autant moins acquis que le chercheur souhaitant partir ne peut prétendre le faire que sur un poste précaire. Le cas de Georges diffère, puisque ses déplacements successifs lui permettent très directement d’accéder à des postes stables, mieux rémunérés. Il s’agit aussi du seul des quatre cas où la décision de départ est suivie par la conjointe, aux dépens de sa propre carrière. Cette décision conjointe n’est pourtant pas réalisée sans coût, puisque Georges précise que le suivi au Canada de sa première partenaire a largement contribué à la rupture du couple. 11.6 Conclusion L’analyse combinée des données des questionnaires et des entretiens permet de mettre en évidence dans quelles mesures « faire un postdoc en Suisse » renvoie à des pratiques, des vécus subjectifs et des réalités multiples. À partir d’une ACM incorporant plusieurs indicateurs traditionnellement associés à l’avancement des carrières, tels que la durée de l’expérience postdoctorale, le 272 lieu, la discipline de soutenance, ou encore le volume de publications, nous formulons une typologie des devenirs professionnels qui s’ouvrent – au cœur du système helvétique ou à ses marges  – aux jeunes docteur·e·s débutant un postdoc en Suisse. Nous traçons ce faisant un « espace des parcours pos- tdoctoraux » sensible à plusieurs propriétés sociales, en particulier relatives à l’intersection du genre et de la socialisation conjugale, et explorons par le biais d’études de cas l’intrication étroite entre trajectoires professionnelles, géographiques et familiales, dans un contexte académique national peu favorable aux stabilisations. Le modèle traditionnel des carrières helvétiques s’hybride ainsi à d’autres formes de trajectoires, en lien avec deux mouvements participant à la reconfiguration des systèmes de carrière : l’internationalisation du paysage académique et sa dé-standardisation. Plus généralement, nos résultats questionnent une certaines doxa : celle d’une nécessaire et inexorable « internationalisation » des marchés du travail et des carrières – particulièrement forte dans les professions intellectuelles et artistiques, où la problématique de l’attraction des « talents » par-delà les frontières nationales est particulièrement structurante pour les employeurs (Menger 2018). Tout comme les artistes, les universitaires sont ainsi souvent présenté·e·s comme des fers de lances et des précurseur·e·s de la mondialisation des flux de travailleur·e. En échos à de précédents travaux sur les professions artistiques en Suisse (Perrenoud et Bataille 2019 ; Bataille et al. 2020), l’analyse des cheminements postdoctoraux montre que les logiques façonnant l’espace national de départ et d’arrivée restent extrêmement structurantes, que ce soit dans, les stratégies, les représentations ou les pratiques des actrices et acteurs. Il apparaît ici que l’ancrage national des parcours individuel constitue une donnée fondamentale pour comprendre les logiques qui orientent et (dé) favorisent – l’accès à des positions professionnelles stables et valorisantes. Nos analyses pointent en creux la nécessaire distance critique à prendre dès lors qu’il s’agit de lier le sujet de la mobilité et de la compétitivité ou de l’excellence scientifique dans le secteur de l’emploi académique. Ainsi, l’analyse des délibérations lors des recrutements académiques dans les uni- versités suisses montre que les usages sociaux et rhétoriques de la mobilité (pour évaluer une candidature, la soutenir ou au contraire pour le dévaluer) sont multiples, nationalement et disciplinairement situés, et en grande par- tie détachées des pratiques concrètes de déplacement dans l’espace (Sautier 2020). En outre, loin d’être toujours synonyme d’une démultiplication des possibles professionnels, réaliser une mobilité à l’étranger génère également des situations de ruptures professionnelles avec les pays d’origine, vis-à-vis desquels les chercheur·e·s se trouvent ‘locked out’ (Pustelnikovaite 2020). Les réflexions développées ici invitent à se saisir de « la mobilité » moins comme un 273 capital en soi, mobilisable dans des transactions pour négocier de meilleures conditions de travail, mais comme une notion pouvant faire l’objet d’usages sociaux mouvants – et dont les possibles bénéfices sont à contextualiser au regard des logiques symboliques structurant les contextes socio-politiques historiquement et géographiquement situés. 11.7 Références bibliographiques Ackers, Louise. 2008. Internationalisation, Mobility and Metrics: A New Form of Indirect Discrimination? Minerva 46(4): 411‒435. Armano, Emiliana et Annalisa Murgia. 2013. The Precariousnesses of Young Knowledge Workers: A Subject-Oriented Approach. Global Discourse, 3(3‒4): 486‒501. Bataille, Pierre. 2016. Qualitative Report on Leaky Pipeline—University of Lausanne. In Bernard Fusulier et Farah Dubois-Shaik (dir.), Deliverable D6.2, Qualitative report on Leaky Pipeline. Trento: University of Trento, 190‒221. Bataille, Pierre, Nicky Le Feuvre, et Sabine Kradolfer. 2017. Should I Stay or Should I Go? The Effects of Precariousness on the Gendered Career Aspi- rations of Postdocs in Switzerland. European Educationnal Research Journal, 16(2‒3): 313‒331. Bataille, Pierre, Johannes M. Hedinger, et Olivier Moeschler. 2020. Visual Artists’ Professional Situations and Trajectories Between Institutions and the Market. In Andrea Glauser et al. (dir.), The Sociology of Arts and Markets, Sociology of the Arts. Cham: Springer International Publishing, 97‒128. Benninghoff, Martin, Gaële Goastellec, et Jean-Philippe Leresche. 2009. L’in- ternational comme ressource cognitive et symbolique : changements dans l’instrumentation de la recherche et de l’enseignement supérieur en Suisse. In Leresche Jean-Philippe, Larédo Philippe, et Weber Karl (dir.), Recherche et Enseignement Supérieur Face à l’Internationalisation. France, Suisse et Union Européenne: Presses Polytechniques Romandes, 235‒255. Bernela, B., & Milard, B. (2016). Co-authorship Network Dynamics and Geo- graphical Trajectories – What Part Does Mobility Play? Bulletin of Sociological Methodology/Bulletin de Méthodologie Sociologique, 131(1), 5–24. Cantwell, Brendan. 2011. Transnational Mobility and International Academic Employment: Gatekeeping in an Academic Competition Arena. Minerva, 49(4): 425‒445. Cantwell, Brendan et Jenny Lee. 2010. Unseen Workers in the Academic Factory: Perceptions of Neoracism Among International Postdocs in the United States and the United Kingdom. Harvard Educational Review, 80(4): 490‒517. Charle, Christophe et Jacques Verger. 2012. Histoire des universités. XIIème-XXIème siècle. Paris: Presses universitaires de France. 274 Colombi, Denis. 2016. Les usages de la mondialisation : mobilité internationale et marchés du travail en France. PhD dissertation, Institut d’Études Politiques, École doctorale de Sciences Po (Paris), France. Darmon, Muriel. 2008. La notion de carrière : un instrument interactionniste d’objectivation. Politix, 82: 149‒167. del Río Carral, Maria et Bernard Fusulier. 2013. Jeunes chercheurs face aux exigences de disponibilité temporelle. Quelles logiques d’agencement entre la vie professionnelle et la vie privée ? Temporalités. Revue de sciences sociales et humaines, 18. Dorenkamp, Isabelle et Stefan Süß. 2017. Work-life Conflict Among Young Academics: Antecedents and Gender Effects. European Journal of Higher Education, 7(4): 402‒423. Dorenkamp, Isabelle et Eva-Ellen Weiß. 2017. What Makes them Leave? A Path Model of Postdocs’ Intentions to Leave Academia. Higher Education, 75: 747‒767. EACEA. 2017. Modernisation of Higher Education in Europe: Academic Staff—2017. Eurydice Report. Luxembourg: Publications Office of the Euro- pean Union. June, 2017, https://op.europa.eu/en/publication-detail/-/publicati on/40f84414-683f-11e7-b2f2-01aa75ed71a1/language-en (23.03.2021). Enders, Jürgen et Christine Musselin. 2008. Back to the Future? The Academic Professions in the 21st Century. Higher Education to 2030, Demography, OECD Publishing, 1: 125‒150, https://www.oecd-ilibrary.org/education/higher-ed- ucation-to-2030-volume-1-demography/back-to-the-future-the-academic- professions-in-the-21st-century_9789264040663-5-en (23.03.2021). Fassa, Farinaz et Sabine Kradolfer. 2013. The Gendering of Excellence Through Quality Criteria: The Case of the Swiss National Science Foundation Profes- sorships in Switzerland. Tertiary Education and Management, 19(3): 189‒204. Felli, Romain, Gaële Goastellec, et Jean-Philippe Leresche. 2007. Existe-t-il un marché du travail académique franco-suisse ? Formation emploi, (4): 49‒64. Fumasoli, Tatiana, Gaële Goastellec, et Barbara M. Kehm (dir.). 2015. Academic Work and Careers in Europe: Trends, Challenges, Perspectives. Dordrecht: Springer. Fusulier, Bernard et Maria del Río Carral. 2012. Chercheur-es sous haute tension ! : vitalité, compétitivité, précarité et (in) compatibilité travail/famille. Louvain-la- Neuve: Presses universitaires de Louvain. Fusulier, Bernard et Farah Dubois-Shaik (dir.). 2016. Deliverable D6.2, Quali- tative report on Leaky Pipeline. University of Trento: Garcia Report. March, 2016, http://garciaproject.eu/wp-content/uploads/2015/11/GARCIA_tem- plate-with-cover_def1.pdf (23.03.2021). Gingras, Yves. 2002. Les formes spécifiques de l’internationalité du champ scienti- fique. Actes de la Recherche en Sciences Sociales, (141‒142): 31‒45, https://www. cairn.info/revue-actes-de-la-recherche-en-sciences-sociales-2002-1-page-31. htm (23.03.2021). 275 Goastellec, Gaële. 2016. La mobilité internationale : une qualité des carrières et des marchés académiques en Europe ? Journal of international Mobility, (4): 171‒188. Goastellec, Gaële. 2017. Internationalization of Higher Education Research and Careers, Europe. In Teixeira Pedro et Shin Jung Cheol (dir.), Encyclopedia of International Higher Education Systems and Institutions. Dordrecht: Springer. Henderson, Emily. 2020. Sticky Care and Conference Travel: Unpacking Care as an Explanatory Factor for Gendered Academic Immobility. Higher Education. Horta, Hugo. 2009. Holding a Post-Doctoral Position before becoming a Faculty Member: Does it Bring Benefits for the Scholarly Enterprise? Higher Education, 58(5): 689‒721. Jöns, Heike. 2011. Transnational Academic Mobility and Gender. Globalisation, Societies and Education, 9(2): 183‒209. Le Feuvre, Nicky. 2015. Contextualizing Women’s Academic Careers in Cross-National Perspective. GARCIA Working Papers No. 3, GARCIA—GA No. 611737. Trento: University of Trento. https://serval.unil.ch/resource/ serval:BIB_07D9E0298FB3.P001/REF (23.03.2021). Le Roux, Brigitte et Henry Rouanet. 2010. Multiple Correspondence Analysis. London: SAGE. Leemann, Regula Julia, Philipp Dubach, et Stefan Boes. 2010. The Leaky Pipeline in the Swiss University System: Identifying Gender Barriers in Postgraduate Education and Networks Using Longitudinal Data. Revue Suisse de Sociologie, 36(2): 299‒323. Menger, Pierre-Michel. 2018. Le talent en débat. Paris: Presses Universitaires de France. Murgia, Annalisa et Barbara Poggio. 2014. At Risk of Deskilling and Trapped by Passion: A Picture of Precarious Highly Educated Young Workers in Italy, Spain and the United Kingdom. In Antonucci Lorenza et Hamilton Myra (dir.), Young People and Social Policy in Europe. London: Palgrace Macmillan, 62‒86. Musselin, Christine. 2017. La grande course des universités. Paris: Presses de Sciences Po. Ortlieb, Renate et Silvana Weiss. 2018. What Makes Academic Careers Less Insecure? The Role of Individual-Level Antecedents. Higher Education, 76: 571‒587. Paye, Simon. 2015. Temporaires, permanents, « vedettes » : la tripartition du salariat académique au Royaume-Uni. Sociologie du travail, 57(2): 175‒199. Perrenoud, Marc et Pierre Bataille. 2019. Vivre de la musique ? Lausanne: Antipodes. Pustelnikovaite, Toma. 2020. Locked Out, Locked In and Stuck: Exploring Migrant Academics’ Experiences of Moving to the UK. Higher Education. Powell, Kendall. 2015. The Future of the Postdoc. Nature News, 520(7546): 144. 276 Renisio, Yann et Rémi Sinthon. 2014. L’analyse des correspondances multiples au service de l’enquête de terrain. Genèses, 97: 109‒125. Rolle, Valérie. 2016. Les corps professionnels en jeu. Genèses, 104: 115‒132. Rossier, Thierry, Marion Beetschen, André Mach, et Felix Bühlmann. 2015. Inter- nationalisation des élites académiques suisses au XXe siècle : convergences et contrastes. Cahiers de la recherche sur l’éducation et les savoirs, (14): 119‒139. Sautier, Marie. 2020. « Recruiting and Assessing Academics Across Geographical Borders ». Annual Meeting of the Society for Social Studies of Science, 18–21 Aug 2020, Czech Technical University, Prague, https://www.easst4s2020prague. org/ (23.03.2021). Sautier, Marie. 2021. (article en préparation). Move or Perish? Sticky Mobilities in the Swiss Academic Context. Higher Education. Schaer, Martine, Janine Dahinden, et Alina Toader. 2017. Transnational Mobil- ity Among Early-Career Academics: Gendered Aspects of Negotiations and Arrangements within Heterosexual Couples. Journal of Ethnic and Migration Studies, 43(8): 1292‒1307. Schultheis, Franz. 2000. Un inconscient universitaire fait homme : le Privatdozent. Actes de la recherche en sciences sociales, 135(135): 58‒62. Shinozaki, Kyoko. 2017. Gender and Citizenship in Academic Career Progression: An Intersectional, Meso-Scale Analysis in German Higher Education Institu- tions. Journal of Ethnic and Migration Studies, 43(8): 1325‒1346. Storme, T., J. R. Faulconbridge, J. V. Beaverstock, B. Derudder, et F. Witlox. 2017. Mobility and Professional Networks in Academia: An Exploration of the Obligations of Presence. Mobilities, 12(3): 405‒424. Theodosiou, Maria, Jean-Philippe Rennard, et Arsia Amir-Aslani. 2012. The Rise of the Professional Master’s Degree: The Answer to the Postdoc/PhD Bubble. Nature Biotechnology, 30(4): 367‒368. Yang, Lijing et Karen L. Webber. 2015. A Decade Beyond the Doctorate: The Influence of a US Postdoctoral Appointment on Faculty Career, Productivity, and Salary. Higher Education, 70(4): 667‒687. Ylijoki, Oili-Helena. 2003. Entangled in Academic Capitalism? A Case-Study on Changing Ideals and Practices of University Research. Higher Education, 45(3): 307‒335. Zippel, Kathrin. 2017. Women in Global Science. Advancing Academic Careers through International Collaboration. Stanford University Press. 277 12 Career Starts of Swiss Sociologists: Labour Market and Working Conditions for Sociologists Rainer Diaz-Bone 12.1 Introduction Sociology is a young scientific discipline. It was Auguste Comte, who coined the term ‘sociology’ in the middle of the 19th century (Comte 2009).1 But it took sociology many more decades to form as a scientific discipline and to institutionalize first chairs and journals. This happened since the turn of the 20th century (starting in the US, Germany, and France). And it took again some decades to implement university study programs in sociology.2 Today, sociology can be regarded as a foundational discipline for the social sciences, providing methodologies and theories to neighbouring disciplines such as political or educational sciences. It has its place in the system of scientific disciplines. With the educational expansion, more and more universities were es- tablished, and the number of academic positions for sociologists increased quickly. For round about half a century now, most trained sociologists do find their job opportunities outside the higher education system. Social sciences in general and sociology, in particular, have become modernizing forces in many social fields and service branches. Sociologists have accessed more and more 1 For him, sociology would be the most advanced discipline, after the sciences would have overcome their theological and metaphysical stages, becoming a ‘positivist’ science in the ‘positivist stage’ of human knowledge (Comte 2009). 2 Although in Switzerland, the University of Geneva was the first to offer a Sociolo- gy PhD program in the 1890s, the university training in Swiss sociology has been established only since the 1950s. The Swiss Sociological Association was founded in 1955 (Eberle 2005; Eberle and Reichle 2018; Holzhauser et al. 2019, 49). See for the history of sociology and its institutions in Switzerland Zürcher (1995) and Holzhauser et al. (2019). 278 job positions for professional activities as planning, evaluation, counselling, research, managing, and governance. There are academic disciplines, for which the state controls access to and performance of professions—as in the case of medical professions, teachers, or lawyers—and some disciplines have invested for decades in politics to establish their discipline as a prestigious profession, too—as economics or psychology have engaged for. In these cases, corresponding professional categories outside the academia are established in long-lasting processes of lobbying for their establishment and as a result of social actors enforcing their social recognition and legitimization.3 For these disciplines, the labour market outside the higher education system is highly visible and, in many cases, also represented in the mass media. Also, some of these academic groups are related to lifestyle images—as medical professionals or law professionals—, which is not to say that the everyday routines and practices or the real lifestyles of these professions are validly known or represented in society. In contrast to these sketched groups, for many scientific disciplines, there is no clearly visible labour market existing, except professions in the higher education system as university professors. Most disciplines in the field of the humanities or disciplines as political sciences or sociology have no visible professional identity and no visible professional career paths. As soon as sociologists leave the higher education system (as teachers and researchers at universities and universities of applied science), step by step, they ‘invisibilize’ themselves as graduates of their discipline. Although they are successful in many different jobs and economic branches, they tend not to be recognized as sociologists, but as possessors of positions in organizations as counsellors, evaluators, market researchers, staff managers and others. The problem, related to this mechanism, is that school leavers will not have an idea about sociologists’ labour market. One can speak of a ‘labour market confusion’, i.e. sociology students, media professionals and also the wider public has no clear idea about sociologists’ work outside of the higher education system (Diaz-Bone and Jann 2019, 1).4 Sociology regards professions and work as basic sociological concepts and explains social structure (and social inequality) mainly with regard to 3 Scholars as Pierre Bourdieu (1984a) or Andrew Abbott (2001) have studied the social conflicts, which arise around professional categories and the division of labour. See Boltanski (1987) for a historical study of the professional category of the ‘cadres’ in France, a group, which did not exist until the middle of the 20th century, whose implementation was first tried with engineers as its representation and which is today the biggest socio-professional category in France (Desrosières and Thévenot 2002). 4 See for more references Diaz-Bone and Jann (2019). 279 the system of professions (Weber 1978; Blau and Duncan 1967; Bourdieu 1984b; Abbott 2001). But so far, there is no strong tradition in the sociology of sociology as profession itself. The aim of this contribution is to track sociologists’ first years of their career outside the academia. Therefore, data from the survey program ‘Absol- ventenstudien Hochschulen’ (Graduate Studies—Higher Education System) will be analysed, which is run by the Swiss Federal Statistical Office (FSO). With the higher graduate studies data, it will be possible to analyse the first five years of career paths of graduates from Swiss universities. Here, the focus is on graduates in sociology, who were employed within the first five years after graduation. This way, the conditions of sociologists’ employment during the beginning of their career can be identified, typical beginnings of career paths identified and compared. Although sociology cannot be grasped as an established professional category, it is possible to approach the different realities of sociological labour by combining variables that represent separate aspects of labour and jobs. Professions are complex phenomena (Abbott 1988; Desrosières and Thévenot 2002).5 To interpret constellations of categories multiple correspondence analysis (MCA) is a standard approach to social positions in contemporary sociology (Bourdieu 1984a; Bourdieu 1984b).6 After this introduction, a short sketch of results from foregoing studies is presented (section 2). Then, the data from the survey program of the FSO and the development of graduates (in sociology, in the social sciences, and for all graduates) for the time span under study are presented. Also, problems and limitations of data quality are considered (section 3). The following part of this contribution focuses on the first years of sociologists’ career. Here, sociologists are compared with the bigger groups of all social scientists and all graduates as to how many job positions they had in the first five years, if they remained in economic sectors (or not), how incomes have developed and what positions they have achieved (section 4). Afterwards, the working conditions of sociologists are analysed. Applying MCA, typical constella- tions of job properties (income, workload, economic sector, position, type of contract) and the main principles (‘axes’), which structure the (sub)field of sociologists’ labour market are identified (section 5). Finally concluding remarks and some outlooks on trends and perspectives of the labour market for sociologists will be given (section 6). 5 See for an analysis of jobs and professions of social scientists Diaz-Bone et al. (2004). 6 Nowadays MCA can be easily proceeded by established statistical data analysis software as R. See Le Roux and Rouanet (2004; 2010) and Husson et al. (2017). 280 12.2 What is Known This analysis continues some foregoing studies, which approach the situation of Swiss graduates in sociology one year after graduation (Diaz-Bone 2019a) and five years after graduation separately (Jann 2019).7 Results show that sociologists successfully enter the labour market after one year although sociologists need half up to two months longer to access a first adequate position (Diaz-Bone 2019a). Their unemployment rates have been above the average at the beginning of the 2000s but have been decreasing down to 2.7 percent in 2017, which is the average over all graduates. Therefore, unem- ployment cannot be regarded as a problem for Swiss graduates in sociology (in the time span under study), neither one year after graduation nor five years after graduation. Compared to all disciplines the proportion of PhDs in sociology is above the average since 2009 and sociologists are slightly more likely to achieve a management position than the average of Swiss graduates (Diaz-Bone 2019a). Compared with other social sciences, psychology, and humanities, the group of graduates in sociology has a little less median income and report slightly less adequacy of the current job to graduate’s qualification (Jann 2019). At the end of the 2010s, Swiss graduates in sociology can be regarded as an academic group, which is successful in regard of its labour market access and positioning.8 12.3 Data and Data Quality The database is the higher graduates study (‘Absolventenbefragung Hoch- schulen’). The target population is all graduates from Swiss universities, Swiss universities of applied sciences and Swiss universities of teacher education. The design is a repeated (two-wave) panel design, which can therefore also be conceived as a trend design. The whole target population of graduates with BA, MA, teacher’s diploma (TeachD) and PhD are surveyed (census).9 Every panel comprises two waves: The first survey is run one year after graduation. 7 These analyses where initiated by the board of the Swiss Sociological Association (SSA) and first results were presented at the sociology congress 2019 at the University of Neuchâtel and later on published in the SSA Bulletin 155. 8 For an analysis of graduates in sociology from University of Geneva see Rosenstein and Widmer (2019) with similar results (as sketched above for all Swiss graduates) for this group. 9 BA grades and MA grades are included since 2007. The FOS has recoded older university diploma and licentiate into MA and diploma of universities of applied sciences as BA. 281 The second wave is run four years after the first interview, which is five years after graduation (only with graduates who responded to the first survey). Every two years, a new panel is proceeded. The Swiss FSO is running the higher graduates study since 1998. The questionnaire was revised in 2003; therefore, comparability is given for the data sets from 2003 onwards; data is available from FSO up to 2017. For this contribution, the surveys (first wave/second wave) from 2003/2007, 2005/2009, 2007/2011, 2009/2013, 2011/2015, 2013/2017 were used. The survey mode is a web-based survey (Federal Statistical Office 2016). The Swiss FSO reports the response rates to be round about 60 percent for the first waves and round about 65 percent for the second waves (Federal Statistical Office 2016).10 The Swiss FSO has provided weights (always for the first and second wave separately) to com- pensate for nonresponse (Federal Statistical Office 2016).11 In regard to data quality one has to expect more non-response of graduates, who switched the town and workplace more often in the course of the first five years. The address and contact management here is presumably more flawed. (Since the calculation of weights focuses the adjustment for disciplines and institution, the bias due to higher job mobility will negatively affect the data quality.) For the analysis in this contribution only those graduates were selected, who left the higher education system after graduation12 and have been employed at both dates of the two interviews. The focus is on employed sociologists and their first five years of labour.13 10 For the included data sets, self-calculated response rates for the first waves range from 53 percent to 64 percent and for the second waves response rates range from 42 percent to 70 percent (decreasing over time). Finally, both waves in each panel represent data in a range of 35 percent to 58 percent of the target population. 11 The principle to calculate weights was to approximate the number sum of cases in the population for disciplines and universities (Federal Statistical Office 2016). Reported results in this contribution are weighted, except for MCA in section 5. 12 Many BA graduates, who were interviewed one year after graduation continued their academic qualification and stayed in the higher education system aiming for a MA degree. A part of them is included in the respectively following panel two years later as MA graduates entering the labour market. 13 This strategy of subsetting ensures that only those graduates are included, who left the university education system after graduation for a professional activity and who did not aim for another academic grade until the date of the second interview. Excluded are those graduates, who started a career but stepped out of employment until the date of the second interview. 282 Where suitable, sociologists will be compared to the whole group of social scientists14 and to the graduates of all disciplines (i.e. the totality of graduates)—this way, the smaller group is always included in the next bigger one.15 In some cases, pooling (accumulation) is applied (also to deal with a small number of cases). In the period studied, sociologists represent less than 1 percent and social scientists up to 10 percent of all graduates.16 Table 1 shows that the absolute number of graduates in sociology tends to be low. Table 1: Percentages of employed graduates by academic degree 2003/ 2005/ 2007/ 2009/ 2011/ 2013/ 2007 2009 2011 2013 2015 2017 Sociology BA 0.00 0.00 0.00 8.33 8.49 13.40 MA 82.86 90.57 89.42 79.17 74.53 61.86 PhD 17.14 9.43 10.58 12.50 16.98 24.74 S 100.00 100.00 100.00 100.00 100.00 100.00 n 70 106 104 120 106 97 Social sciences BA 0.00 2.52 3.06 8.18 11.85 15.54 MA 91.08 84.94 87.44 78.85 74.29 68.82 PhD 8.92 12.54 9.50 12.97 13.86 15.64 S 100.00 100.00 100.00 100.00 100.00 100.00 n 1 188 1 428 1 537 1 858 1 696 1 886 All disciplines BA 30.64 31.52 36.78 34.91 33.04 34.13 MA 57.91 48.73 36.31 37.07 38.13 38.90 TeachD 0.00 4.65 11.96 13.93 14.26 13.71 PhD 11.45 15.10 14.95 14.09 14.58 13.26 S 100.00 100.00 100.00 100.00 100.00 100.00 n 11 726 15 202 17 228 18 998 20 017 22 471 14 The category ‘social sciences’ of FSO includes these categories: human geography, media sciences, political sciences, psychology, pedagogy, sociology, social work, special education, interdisciplinary/other social sciences. 15 This strategy of using these three categories is applied here to use the bigger group as reference (using average) and to enhance comparability with the analysis presented in Diaz-Bone (2019b). 16 See also Diaz-Bone (2019a), Table 1. 283 BA graduates in sociology are missing in the first three panels and there are also no BA graduates in social sciences in the first panel.17 Table 1 also demonstrates that the number of graduates in sociology does not follow the general increasing trend of graduates when compared to social sciences and to all disciplines, which doubled almost in the time span under study. Instead, the number of graduates in sociology has declined from 2009 onwards (see the upper pane of Table 1). 12.4 Sociologists’ Career Starts in Comparison Some indicators can achieve first insights about the dynamics of the start of the career. Table 2 shows the pooled panel data, split up for the three groups and the different grades. The second column gives the average number of the contemporary job at the time of the second interview for every group. For every academic grade, the averages for sociologists are higher if compared to the bigger groups. Sociologists have had more job switches in the course of the first five years than the average in the social sciences and the average of all sciences. In the social sciences job mobility is higher also for BA graduates as for MA graduates. PhDs have the lowest average, they tend to have their second job, when interviewed five years after graduation. Sociologists’ higher average numbers in jobs during the first five years can indicate sociologists’ bigger difficulties to find a position adequate and/or interesting for sociologists. Also, the reported higher average number of months for sociologists to access a first adequate position indicates this (Diaz-Bone 2019a). However, the data can be interpreted in another way: sociologists’ qualification can be related to a broader range of tasks, too. It might be easier for sociologists to switch jobs, therefore. The next two columns represent the average annual income five years after graduation and the percentage increase of income between the two interviews (i.e. an increase of income in the four years before the second interview). When compared to other European countries, salaries in Switzerland are known to be high. But MA graduates and much more BA graduates in social sciences (including sociology) tend to have a lower income than the corresponding averages of all disciplines.18 It is remarkable that PhDs in sociology tend to have annual salaries considerably above the average of all PhD graduates and also above the average of social science PhDs. The figures in the third column 17 The reason for this is the transformation of study programs and their final grades from diploma to BA and MA starting only in the early 2000s. 18 Jann (2019) has argued that there is no evidence for a significant gender pay gap in the social sciences. 284 Table 2: Job number, income, and position Academic degree Job number Annual Increase of Management Higher (mean) income income (%) E ducation/ (mean in (%) Research 1000 CHF) (%) Sociology BA 2.9 83.5 12.78 46.88 0.00 MA 2.7 93.2 37.37 44.26 6.89 PhD 2.4 119.2 17.08 43.33 41.11 Social sciences BA 2.7 82.6 22.70 25.07 2.05 MA 2.6 91.3 29.28 33.26 5.65 PhD 2.3 109.9 17.11 39.18 34.86 All disciplines BA 2.1 92.5 22.00 43.65 1.06 MA 2.4 96.5 33.84 36.30 5.52 TeachD 2.0 91.9 9.57 4.12 0.37 PhD 2.2 108.8 22.98 43.30 21.63 Note: In Table 2, annual income is calculated as standardized average annual gross income five years after graduation. Only graduates with at least 90 percent workload at the second interview are included in the calculation. The increase of income is calculated as the weighted mean of individual percentage increase. Only graduates with at least 90 percent workload at time of the first and second interview are included in the calculation. For the last column, the higher education category includes professors, research staff, and lecturers. In this category PhD students are excluded, who are engaged in research projects. In Switzerland, PhD students are funded with only 50 percent workload by the Swiss National Science Foundation (SNSF). of Table 2 suggest that for BA graduates in the social sciences (including sociology), an additional MA grade is a good investment and especially for sociologists a PhD is a worthwhile grade. But it is also the MA graduates across all three groups (compared here), who have experienced the biggest percentage increase in income during the four last years working (instead of aiming for a PhD). When considering income, it is advisable for social scientists not to enter the labour market with a BA grade. When looking on the average percentage increase of the annual income (in the course of the last four years), BA graduates and PhDs in sociology do not perform as well as the averages of all graduates. MA graduates in sociology do perform better than every other group in Table 2. The reason for this is not clear here and needs further investigation. In sum, the situation in terms of income is good for social scientists in general and sociologists in particular. (Table 2 confirms the results in Jann 2019). The last two columns in Table 2 again make evident that sociologists tend to be more likely to achieve man- 285 agement positions. Especially PhDs in sociology are successful in entering the higher education system as teachers, lecturers, and researchers. What about the job mobility of employed sociologists concerning the public sector, non-government organizations (NGOs), and companies (for-profit sector)? Table 3 represents again pooled data for sociologists in comparison to all social sciences and to all disciplines. In the upper part of Table 3, one can see which percentage of employed sociologists in one sector stayed there or switched until the second interview four years later. Table 3: Percentages of employment sectors First wave Second wave S n % of total Public NGO Company (%) (%) (%) Sociology Public 84.51 6.34 9.15 100.00 284 49.91 NGO 19.78 60.44 19.78 100.00 91 15.99 Company 15.98 7.22 76.80 100.00 194 34.09 S 50.79 15.29 33.92 100.00 569 100.00 Social sciences Public 81.52 8.09 10.39 100.00 5 033 55.34 NGO 35.72 44.63 19.65 100.00 1 257 13.82 Company 18.87 7.20 73.93 100.00 2 804 30.83 S 55.87 12.87 31.26 100.00 9 094 100.00 All disciplines Public 81.51 5.11 13.38 100.00 45 986 45.32 NGO 34.44 42.60 22.96 100.00 6 955 6.85 Company 14.94 2.99 82.07 100.00 48 535 47.83 S 46.45 6.66 46.89 100.00 101 476 100.00 Note:: Categories in the rows are employment sectors at time of first interview (first wave), colum categories are employment sectors at time of second interview (second wave); for categories see text above. It turns out that almost half of the sociologists work in the public sector at both interviews (49.91 % and 50.79 %). The big majority of sociologists, who worked in the public sector early on (84.51 %) stayed there. The inclination to stick to the public sector is shared with social science graduates and in fact, all graduates in general, where the average percentages are of the same level. Also, sociologists tend to stick to the company sector, but to a lesser extent than to the public sector. In comparison to the whole group of social scientists and to all graduates, sociologists have a stronger affinity to NGOs. 286 They work more often in NGOs at both interview dates (15.99 % and 15.25 %) than the two embedding larger groups. Sociologists have a much higher percentage of employees, who stick to this sector (60.44 %) than the social scientists (44.63 %) or graduates of all disciplines have (42.60 %). In- terestingly, sociologists have a higher inclination to stick to one sector (which is equivalent to a lower mobility between sectors) than the average of social scientists. (All percentages in the first diagonal are higher for sociologists than for social scientists.) When reminding the higher number of jobs, sociologists pass through (Table 2), they presumably change jobs more likely in the same sector. It is important to recognize that the non-profit sector (public sector and NGOs) is most important for all social scientists as the labour market. In this regard, social scientists deviate from the average of all disciplines. 12.5 Sociologists’ Labour Market and Working Conditions Sociology is not a profession, as lawyer or medical doctor are. To inspect classifications of occupations and their occupational categories, therefore, is a strategy, which will not lead any further as identifying only a few established job categories to which some sociologists oriented their career. But in many cases, the occupational category with the highest count will be ‘not classifi- able’, as it is the case for sociologists (see Diaz-Bone 2019a). Therefore, sociologists’ professional position and working conditions are analysed not by relying on conventional professional categories, which are not valid categories for many sociologists’ professional activities and posi- tions. Instead, the analysis is based on aspects, which can be reported reliably and which are regarded as valid indicators also for sociologists working in different jobs. To analyse the correlations between properties of jobs and positions, which sociologists have, a multiple correspondence analysis (MCA) is calculated for the data of the second wave, representing the employment situation five years after graduation. In fact, the resulting graphical outputs of MCA represent the subfield of sociologists’ labour market, articulating its main principles in Switzerland for the time period under study. 12.5.1 Included Cases and Used Variables The multiple correspondence analysis (MCA) is calculated with pooled data. Only sociologists (included in one of the six surveys from 2007 to 2017) with 287 an employment at the date of the second interview are included.19 They can be regarded as part of the labour force in Switzerland, while others stepped out. In this MCA 317 sociologists are included in the data set. Before introducing the results of MCA procedures, the employed variables and the frequencies of the involved categories will be presented. For the MCA, the position in an organization (professional status), the workload (volume of work or pensum), the kind of sector to which the or- ganization belongs, the annual income, and the kind of contract are used. To locate afterwards different groups in the labour market, gender and academic degree are included later on as supplementary (or passive) variables.20 › Sector separates the already applied categorization for economic sectors (see Table 3), which are public sector, NGOs, and companies, also an else category is included. › Position has four categories: management, employee, researcher or teacher in higher education, and an else category.21 › Workload is categorized into three categories: less than 50 percent, up to 90 percent and full-time job. › Annual income is categorized into less than 70 thousand Swiss francs, up to 90 thousand CHF, up to 110 thousand CHF and 110 thousand CHF or more. › Contract is dichotomous and differentiates fixed-term employment contract and unlimited working contract. Table 4 presents the univariate distributions of the five involved active variables. 19 Analysis is proceeded without weighting of cases. Cases with missing values are omitted. 20 The categories of supplementary variables do not influence the calculation of the axes and the distribution of the categories of the first set of variables. The axes are calculated on the basis of the active variables first. Therefore, the interpretation of the axes and the planes refers to the categories of this active variables. The labour market as a multidimensional ‘space’, constructed by the axes of MCA therefore will be interpreted in terms of working conditions and labour market aspects. For a better understanding, where social groups in terms of socio-demographics (gender and degree in higher eduction) are located, the categories of the supplementary are included in Figure 1 and Figure 2. 21 No sociologist in the data set has worked self-employed, therefore this category is not included in the MCA. 288 Table 4: Active variables n % n % Sector Position public sector 159 50.16 management 120 37.85 NGO 47 14.83 employee 128 40.38 private sector 109 34.38 researcher, teacher 48 15.14 else 2 0.63 else 21 6.62 S 317 100.00 S 317 100.00 Workload Annual income (in CHF 1000) < 50 % 22 6.94 < 70T 53 16.72 50 to < 90 % 134 42.27 70 to <90T 123 38.80 90–100 % 161 50.79 90 to <110T 63 19.87 ≥110T 78 24.61 S 317 100.00 S 317 100.00 Contract unlimited 236 74.45 fixed-term 81 25.55 S 317 100.00 Table 5 presents the bivariate distribution of the two supplementary variables and demonstrates that the majority of working sociologists has achieved a MA degree (last two columns in Table 5). This also has been the case for all graduates in sociology (Table 1). In the data set round about two-thirds of sociologists are women (63.09 %) and one third are men (36.91 %). Table 5: Supplementary variables Female Male All n % n % S % BA 15 7.50 5 4.27 20 6.31 MA 170 85.00 94 80.34 264 83.28 PhD 15 7.50 18 15.38 33 10.41 S 200 100.00 117 100.00 317 100.00 Inspecting the column percentages (third and fifth column), it can be seen that men with a PhD prove to be slightly overrepresented, while women with a BA degree are slightly overrepresented. Men tend to have higher degrees, i.e. women tend to leave universities earlier. 289 12.5.2 Multiple Correspondence Analysis (MCA) An advantage of MCA is its ability to visualize the pattern of correlations between categories. This facilitates the identification of categories which are typically combined on the level of individuals. MCA plots categories close to each other, which are higher correlated. (And vice versa, categories, which are plotted remote to each other are slightly or not correlated.) Also, MCA is able to calculate dimensions as independent axes. Similar to principal component analysis (PCA) as a multivariate method to extract and analyse factors, MCA extracts axes, which can be regarded as major latent and more general prin- ciples, which separate categories to opposed clusters of categories (Le Roux and Rouanet 2004; Le Roux and Rouanet 2010; Diaz-Bone 2019b).22 This way, the working conditions of the early years of sociologists’ career can be interpreted based on groups of categories and their oppositions (which are to be explained by referring to the axes). Following Pierre Bourdieu’s way to apply MCA one can conceive the resulting plots as representations of field or subfields (see Bourdieu 1984a; Bourdieu 1984b). In this study, a subfield of the Swiss academic workforce is approached by MCA and it can be conceived also as a representation of sociologists’ labour market (Diaz-Bone et al. 2004). It has to be reminded to the fact that the data set only represents the younger part of working sociol- ogists (five years after graduation) and does not entail sociologists in their later steps of their professional careers. In this analysis, three axes are included. The first three axes gather in sum almost 40 percent of the statistical information of the data, which is called (total) inertia. This is a good result for a projection of the categories on three dimensions.23 Percentual contributions to the total inertia are 17.82 percent for the first axis, 11.18 percent for the second axis and 10.29 percent for the third axis.24 22 As in PCA, the axes in MCA can be conceived to be the (latent) factors, explaining the correlation pattern between the categories. While PCA starts with a set of metric variables, MCA calculates the axes starting with categorical variables. Another dif- ference to PCA is the absent practice of rotation of axes in MCA (see Husson et al. 2017, although rotation of axes in MCA is possible). 23 In sum, there are twelve axes existing. The number of axes equals the difference of active variables (5) and the sum of their categories (17). Because the share of inertia is step by step declining from the former axes to the latter ones, the remaining axes after axis 3 are not reported here, which is an usual analytical practice in MCA. 24 The share of percentual contribution of axes declines with the descending order of extraction. In this case, one can identify a bigger gap (of contribution to the total inertia) between axis 1 and the following axes. This provides evidence for axis 1 to 290 MCA’s most important outputs are plots, which employ pairwise two axes.25 Figure 1 shows the resulting MCA plot of the categories in the plane built by the first two axes. Figure 2 shows the MCA plot of the categories in the plane built by the first axis and the third axis.26 Three constellations of categories can be identified. In the upper left of Figure 1, the private sector, management positions, high salaries, and full workloads are gathered. In the upper right, public sector, working positions in the higher education system, fixed-term, lower salaries, and workload are clustered. In the lower part of Figure 1, employees, medium salaries, and jobs in NGOs are grouped. Figure 1: Labour market and working conditions (axes 1 and 2) ≥110T 1.0 PhD resarcher/ manager teacher ≤ 50% in higher 0.5 < 70T education 90–100% BA other position male public fixed-term private sector sector employment 0.0 contract unlimited female working MA contract 90 to < 110T –0.5 NGO 50 to < 90% 70 to < 90T employee –1.0 0.0 1.0 2.0 Dim 1 (17.82%) Note Active variables: sector; position; workload; annual income (in 1000 CHF); contract. Supplementary variables: academic degree; gender represent the most substantial principle for differences of categories. The percentual contribution of the following axes are below 10 percent. 25 The category ‘S_else’ is not plotted in Figure 1 and Figure 2. This category represents only 2 individuals and is located far away from the other categories. (Coordinates are –0.46 for axis 1, –3.12 for axis 2 and 3.80 for axis 3.) 26 The MCA was proceeded with the R package FactoMineR (Husson et al. 2017). See http://factominer.free.fr/ (accessed March 25, 2021). Dim 2 (11.18%) 291 In Figure 1, the axes visualize the main cleavages of sociologists’ working conditions. The interpretation of axes relies on the categories, which have contributions to the axes’ inertias above the average (Le Roux and Rouanet 2010).27 The horizontal axis 1 is mainly linked to the opposition between the private sector, management positions, and unlimited job contracts on the left side and the public sector, having a position for higher education or research, and fixed-term job contracts on the right side. Axis 1 presents mainly the opposition of working conditions for sociologists between the private sector (left side) and the public sector (right side). Axis 2 represents the opposition between high salaries (≥ 110T), full workloads (90‒100 %), and management positions on the top and em- ployees’ positions, medium salaries (70 to < 90T), and reduced workloads (50 to < 90 %) at the bottom. Axis 2 mainly presents the intensity of profes- sional commitment and success. The supplementary categories confirm the result of higher university grades correlating with higher salaries. The location of the women category in the lower right quadrant displays women not to be proportionally equipped with management positions, full workload, and high salaries as male sociologists are. Figure 2 shows the plane, which is constructed by axis 1 and axis 3. The axis 3 opposes less attractive working conditions in the non-public sector (private sector, other sectors, annual income below 70 000 CHF [≤ 70T], workload between 90 and 100 %, other position)28 to more attractive working conditions in the NGO sector (NGO, income between 90 and 110 000 CHF [90 to < 110T], workload between 50 and 90 %,manager). Women and BA graduates are more closely related to the less attractive working conditions in the private sector, while men, MA graduates, and PhDs are more closely linked to the more attractive workings conditions in the NGO sector. In sum, the results of MCA show the variations of sociologists’ working conditions and the cleavages of their labour market. It is the first axis, which represents the most important opposition in the labour market of sociologists and which opposes two sectors—the private sector and the public sector—, which are linked to different constellations of working conditions. When bringing in the categories of the two supplementary variables, it can be seen that BA graduates and women are more likely to experience lesser favourable 27 Because of the number of 17 categories, the average percentage of contribution to an axis is 5.88 percent. See Table 6 in appendix for the contributions of categories to the first three axes. 28 The category other sectors lies outside the frame of Figure 2 (coordinate for axis 3 is 3.80). 292 working conditions than PhDs and men (especially in terms of income, type of contract, and position). Figure 2: Labour market and working conditions (axes 1 and 3) other position < 70T 1.0 private sector BA 0.5 ≤ 50% 90–100% fixed-term employee employment 70 to < 90T contract 0.0 unlimited female working MA contract male public sector resarcher/ –0.5 manager ≥110T 50 to <90% teacher PhD in higher education NGO 90 to < 110T –1.0 –1.0 0.0 1.0 2.0 Dim 1 (17.82%) Note Active variables: sector; position; workload; annual income (in 1000 CHF); contract. Supplementary variables: academic degree; gender 12.6 Perspectives As the preceding sections have demonstrated, sociologists don’t have a clear cut, visible labour market segment and there are no prepared and state-gov- erned career paths established. But sociologists are nevertheless successful in entering the labour market. The trend is positive, when regarding decreasing unemployment rates and higher salaries for PhDs (Diaz-Bone 2019a). One can speak of a ‘labour market paradox’ for sociologists, because these graduates are not visible as a labour market category on one side, but prove Dim 3 (11.18%) 293 to be very successful (even outside the higher education system) on the other side. Representatives of professional academic organizations of sociologists can point to the role, sociology plays as a foundational discipline in the social sciences, providing methods and theories for other disciplines. But it is an open question, what exactly is the explanation for sociologists’ success outside the universities and social research, when graduates cannot rely on an established image of sociologists and their labour market-related skills and competences. It seems a nearby assumption to regard the methodological and conceptual skills of graduated sociologists to be relevant for knowledge production, ‘social engineering’, and coordination of complex social processes outside the academia. Sociology in Switzerland is a small but established discipline, although it has not profited from the general growth of students and graduates in recent years (Diaz-Bone 2019a). One reason for this stagnation of gradu- ates in sociology is the still strong orientation towards science and higher education itself and many sociologists regard ‘science as a vocation’ as Max Weber already has discussed it in a lecture given in 1918 (Weber 2004). Such a strong academic orientation contradicts in a way the appreciation in the Swiss public for more applied academic professions (as engineers, lawyers, teachers, medical doctors, managers, and so on). The Covid-19 pandemic will trouble all graduates who try to enter the labour market. The near future will bring in longer time spans between graduation and recruitment. For sociologists, the negative effects of crises seem to be stronger and longer-lasting than for the social sciences and for all disciplines on average, as has been shown for the financial crises in 2007/2008 (Diaz-Bone 2019a, 8). More severe, the Covid-19 pandemic will also destroy jobs in different industries, but yet the effects for sociologists’ labour market cannot be identified more clearly. Sociologists’ work and sociologists’ labour markets will change in the future—evidently—as they have always changed in the past. But some of today’s and (nearby) future changes are more disruptive. It is not possible to seriously offer predictions, but some trends and perspectives can be sketched. The future of sociologists’ work will be massively influenced by external meg- atrends and already powerful factors, which are trends global in character as digitalization, the globalization of the division of labour, climate change, rising populism, the refugee crisis, or ageing societies (see OECD 2019a; OECD 2019b). These megatrends or factors will generate new sociological debates and inspire sociological research, but will have also influences on labour con- ditions, jobs, and labour market categories. Existing labour market categories always have changed, but these changes will accelerate the evolution of labour 294 markets and many labour market categories will become fuzzier. Sociology as a discipline and as a profession has to adopt to these changes. On one side, more fuzzy labour market categories will fit to sociologists’ existing recogni- tion of labour markets, because many sociologists do not encounter visible jobs explicitly devoted to graduates in sociology (except in the universities and some research institutions). Instead, they have (and had) to apply for open job positions, which are not explicitly devoted to graduates in sociology (or to graduates of one discipline at all). But on the other side graduates of more distant disciplines (as natural sciences or computer sciences) can and do enter fields, where sociologists (and other social scientists) before have been strongly represented as branches like social research or market research. An example is the emergence of data analysts (a truly fuzzy category), whose professional identity is not linked to one discipline and who take over more and more tasks as market research and social research. These are domains which were formerly occupied by sociologists, economists and other social sciences (Davenport and Patil 2012; Harris et al. 2013). Data analysts have to integrate different knowledge areas. Besides data analysis (statistics), it is also substantial domain expertise, which is needed (Schutt and O’Neil 2014). The big Internet companies (as Google), urban startups, and new technolo- gy-based companies have become attractive employers for many graduates. But digitalization and technologies for data production, storage, and analysis move in almost all sectors, and graduates of all disciplines are more and more faced with the expectation of mastering digital skills, code programming and analysis of new forms of data. Sociologists could bring in both, but so far the majority is reluctant to regard advanced digital skills as a necessary major new competence—which has to be combined with substantial sociological knowledge and well-established social research methods. 12.7 References Abbott, Andrew. 1988. The System of Professions. Chicago: University of Chicago Press. Abbott, Andrew. 2001. Chaos of Disciplines. Chicago: Chicago University Press. Blau, Peter and Otis Duncan. 1967. The American Occupational Structure. New York: Free Press. Boltanski, Luc. 1987. The Making of a Class. Cadres in French Society. Cambridge: Cambridge University Press. Bourdieu, Pierre. 1984a. Homo Academicus. Stanford: Stanford University Press. Bourdieu, Pierre. 1984b. Distinction. A Social Critique of the Judgement of Taste. Cambridge: Harvard University Press. 295 Comte, Auguste. 2009. The Positive Philosophy of Auguste Comte. Cambridge: Cambridge University Press. Davenport, Thomas H. and Dhanurjay Patil. 2012. Data Scientist. The Sexiest Job of the 21st Century. Harvard Business Review, October: 70‒76. Desrosières, Alain and Laurent Thévenot. 2002. Les catégories socioprofessionnelles. 5th ed. Paris: La Découverte. Diaz-Bone, Rainer. 2019a. Entering the Job Market for Swiss Sociologists. Bul- letin of the Swiss Sociological Association, 155: 4‒10. https://www.sgs-sss.ch/ en/the-association/bulletin/ (accessed January 30, 2021) Diaz-Bone, Rainer. 2019b. Statistik für Soziologen. 5th ed. Konstanz: UVK. Diaz-Bone, Rainer, Ulf Glöckner, and Anne-Catherine Küffer. 2004. Berufliche Situation und Tätigkeitsfeld von Sozialwissenschaftlern. Eine Analyse mit dem Mikrozensus 2000. Sozialwissenschaften und Berufspraxis, 27(2): 171‒184. Diaz-Bone, Rainer and Ben Jann. 2019. Editorial - Professional Career Paths and the Labor Market for Sociologists. Bulletin of the Swiss Sociological Associa- tion, 155: 1‒3. https://www.sgs-sss.ch/en/the-association/bulletin/ (accessed January 30, 2021) Eberle, Thomas. 2005. 50 Jahre Schweizerische Gesellschaft für Soziologie – 50 ans Société suisse de sociologie. Zürich: Seismo. Eberle, Thomas and Niklaus Reichle. 2018. Die Schweizerische Gesellschaft für Soziologie (SGS). In Stephan Moebius and Andrea Ploder (Eds.), Handbuch Geschichte der deutschsprachigen Soziologie, Band 1: Geschichte der Soziologie im deutschsprachigen Raum. Wiesbaden: Springer VS. Federal Statistical Office. 2016. Survey of higher education graduates (EHA)). Neuchâtel: Swiss Federal Statistical Office. https://www.bfs.admin.ch/bfsstatic/ dam/assets/7753/master (accessed March 25, 2021) Harris, Harlan D., Sean Patrick Murphy, and Marck Vaisman. 2013. Analyzing the Analyzers. An Introspective Survey of Data Scientists and their Work. Sebas- topol: O’Reilly. Holzhauser, Nicole, Andrea Ploder, Stephan Moebius, and Oliver Römer. 2019. Handbuch Geschichte der deutschsprachigen Soziologie. Band 3: Zeittafel. Wies- baden: Springer VS. Husson, François, Sébastien Lê, and Jérôme Pagès. 2017. Exploratory Multivariate Analysis by Example Using R. Abingdon: Chapman and Hall. Jann, Ben. 2019. Early-Career Labor Market Situation of Swiss Sociology Graduates at the Master’s Level. Bulletin of the Swiss Sociological Association, 155: 11‒20. https://www.sgs-sss.ch/en/the-association/bulletin/ (accessed January 30, 2021) Le Roux, Brigitte and Henry Rouanet. 2004. Geometrical Data Analysis. Dord- recht: Kluwer Academics. Le Roux, Brigitte and Henry Rouanet. 2010. Multiple Correspondence Analysis. Thousand Oaks: Sage. 296 OECD. 2019a. The Future of Work. OECD Employment Outlook 2019. Paris: OECD Publishing. OECD. 2019b. OECD Skills Outlook 2019. Thriving in a Digital World. Paris: OECD Publishing. Rosenstein, Emilie and Eric Widmer. 2019. Career Paths of Sociologists from the University of Geneva to the Labour Market. Recent Trends and Future Challenges. Bulletin of the Swiss Sociological Association, 155: 21‒26. https:// www.sgs-sss.ch/en/the-association/bulletin/ (accessed January 30, 2021) Schutt, Rachel and Cathy O’Neil. 2014. Doing Data Science. Sebastopol: O’Reilly. Weber, Max. 1978. Economy and Society. Berkeley: University of California Press. Weber, Max. 2004. The Vocation Lectures. Indianapolis: Hackett Publishing. Zürcher, Markus. 1995. Unterbrochene Traditionen. Die Anfänge der Soziologe in der Schweiz. Zürich: Chronos. 297 Appendix 1 For the interpretation of axes (see section 5), the contributions of categories to axes inertia (variance) are necessary to identify categories which have high contributions to the inertia of axes. Table 6: Contributions of active categories to total inertia of first three axes (in percent) Categories Axis 1 Axis 2 Axis 3 Position management 7.3038 12.5216 6.0051 employee 0.9969 19.7728 3.2404 researcher or teacher in higher education 20.6306 1.6065 0.9941 other position 4.3484 0.3225 8.6753 Contract unlimited working contract 8.5925 0.4075 0.2205 fixed-term employment contract 24.7297 1.1728 0.6346 Sector public sector 11.2776 0.3808 1.4958 NGO 1.5600 2.4609 6.6732 companies 10.1938 0.3454 7.9904 other sector 0.0643 4.6459 7.2483 Workload less than 50 % 1.5520 1.5785 1.9221 50 % up to 90 % 2.6368 9.1783 11.2135 full-time job (90 % up to 100 ) 3.7478 5.2548 6.4244 Annual income less than 70 000 CHF 1.0496 3.4953 17.7336 70 000 CHF up to 90 000 CHF 0.0825 14.6008 2.0865 90 000 CHF up to 110 000 CHF 0.0147 1.9694 13.2589 110 000 CHF or more 1.2189 20.2862 4.1832 S 100.0000 100.0000 100.0000 299 About the Authors Kushtrim Adili is a political scientist and scientific collaborator at the Zurich University of Applied Sciences’ School of Social Work in Switzerland. He is interested in studying the links between migration and social inequality. Philip Balsiger is assistant professor of sociology at the University of Neuchâ- tel, Switzerland. He is interested in the study of social movements, markets, and the transformations of the digital economy. He has recently co-edited the book The Contested Moralities of Markets (Emerald Publishing, 2019). Pierre Bataille holds a PhD in sociology from the University of Lausanne. After being postdoc at the Université Libre de Bruxelles, he is now Maître de conférences at the Université Grenoble-Alpes, conducting research on inequalities in artistic and academic careers in a comparative perspective. Camille Budon is a MA student at the University of Neuchâtel, Switzerland. Her studies focus on volunteering, citizenship, and firefighting. As part of the teaching staff, she supervises BA students and works at the School of Health Sciences Fribourg in a research unit specialized on ageing. Jacinto Cuvi is visiting fellow at the London School of Economics. He holds a PhD in sociology from the University of Texas at Austin. His research fo- cuses on informal labor in the urban South. His work has been published in such outlets as Social Problems, Qualitative Sociology, and Theory and Society. Piergiorgio Degli Esposti is an associate professor at the University of Bologna. His research focuses on digital consumer behaviors and prosumption. His most recent article is The increasing centrality of prosumption in the digital capitalist economy (OJS Austrian Journal of Sociology, 2020). Rainer Diaz-Bone is professor of sociology at the University of Lucerne, Switzerland. His research interests include research methods, social statistics and economic sociology. His most recent book publication is Die “Econo- mie des conventions”. Grundlagen und Entwicklungen der neuen französischen Wirtschaftssoziologie (Springer VS, 2018, 2nd ed). Ulrich Dolata is professor of organizational sociology and innovation studies at the University of Stuttgart, Germany. His most recent book publication is Collectivity and Power on the Internet. A Sociological Perspective (Springer, 2018; with Jan-Felix Schrape). 300 Rainer Gabriel is a lecturer and researcher at the Zurich University of Applied Sciences’ School of Social Work in Switzerland. His research focuses on age- ing and the life course, with particular emphasis on poverty, inequalities in health, and the non take-up of social welfare benefits. Eltje Gajewski is a research associate at the Institute of Sociology, University of Duisburg-Essen, Germany. Her research interests include digital platforms, digitalisation of work, ecological entrepreneurship, and sustainable work. Anna Gonon is research associate at the School of Social Work/FHNW. Her research interests include the sociology of employment and labor market integration. Her recent publication is Legitime Leiden. Die Rechtfertigung von eingeschränktem Arbeitsvermögen und betrieblichen Reintegrationsmaßnahmen (Sozialer Fortschritt, 2021, 70/3: 131–148). Monika Götzö is professor and head of the Institute of Diversity and Social Integration at the Zurich University of Applied Sciences’ School of Social Work in Switzerland. Her research focuses on vulnerability and social integration, combining a cultural studies approach with transdisciplinary stakeholder- involvement in various fields of social work. Anne Jourdain is associate professor of sociology at the Paris-Dauphine University, France. Her research interests include digital work, gender, and career change. She recently co-authored The Social Meaning of Extra Money: Capitalism and the Commodification of Domestic and Leisure Activities (Palgrave Macmillan, 2019). Sonja Kubat is a political scientist and scientific collaborator at the Zurich University of Applied Sciences’ School of Social Work in Switzerland. Her research interest includes urban development, poverty, and social policy. Dominique Méda is professor of sociology at the University Paris Dauphine- PSL, France. Her research interests include relationship to work, social policies, wealth indicators. Her most recent book publication is Manifeste Travail. Démocratiser. Démarchandiser. Dépolluer (Le Seuil, 2020; with Julie Battilana and Isabelle Ferreras). Eva Nadai is professor at the School of Social Work/FHNW. Her research focuses on social policy, welfare institutions and the labour market. Her recent book is Werten und Verwerten. Konventionen der Beschäftigung von Menschen mit Behinderungen in Wirtschaft und Wohlfahrtstaat (Springer VS, 2019; with Alan Canonica, Anna Gonon, Fabienne Rotzetter, and Martin Lengwiler). 301 Mihaela Nedelcu is titular professor of sociology at the University of Neuchâtel, Switzerland. Her main research topics concern migration processes and the social impact of digital technologies. She recently co-edited the special issue Precarious Migrants, Migration Regimes and Digital Technologies (Journal of Ethnic and Migration Studies, 2020). Michel Oris is professor of demography and socioeconomics at the University of Geneva. His research focuses on inequality, with particular emphasis on the living and health conditions of older adults and on vulnerability along the life course. Céline Perrin, a sociologist, is an associate member of the Laboratoire de recherche santé social (Laress) of the Haute école de travail social et de la santé (HETSL/HES-SO) in Lausanne. Her research areas include disability studies, gender studies, and intersectional theory, with particular interest in gender and sexuality. Monika Piecek is a doctoral candidate in the Institute of Gender Studies at the University of Geneva. Her doctoral research focuses on othering processes in Swiss disability insurance drawing on insights from critical disability studies and intersectional theory. Isabelle Probst is associate professor of social psychology at the Haute école de Santé Vaud (HESAV/HES-SO) in Lausanne. Her research interests include disability, experience with social insurance, and various occupational health issues. George Ritzer is a distinguished Emeritus professor at the University of Mary- land. His work focuses on consumer culture, prosumption, globalization, and McDonaldization. His last book is The McDonaldization of Society: Into the Digital Age (Sage Publication, 2018). Marie Sautier is a PhD Candidate at the University of Lausanne and Sci- ences Po Paris. She works on recruitment practices in academia and the internationalisation of the Higher Education system. She has published on the intersections of gender, academic mobility, and precarity. Christian Suter is professor of sociology at the University of Neuchâtel, Switzerland. His research interests include social and economic inequalities, globalization, social indicators, and quality of life. His most recent book publication is The Middle Class in World Society (Routledge, 2020). 302 Jean-Pierre Tabin is full professor of social policy at the Haute école de travail social et de la santé (HES-SO). His research interests include (dis)ability, poverty policies, unemployment insurance, social assistance, homelessness. His most recent book is Repenser la normalité. Perspectives critiques sur le handicap (Le Bord de l’Eau, 2019; with Monika Piecek, Céline Perrin, and Isabelle Probst). Chris Tilly is professor of urban planning and sociology at the University of California Los Angeles. His research examines job quality, inequality, and grassroots collective action. His most recent book is Where Bad Jobs are Better: Retail Jobs across Countries and Companies (Russell Sage, 2017). The Future of Work is a collective, interdisciplinary effort to grasp the trends that were reshaping the world of work before the Covid-19 pandemic along with its effects on the workers and labour markets. Written in the imme - diate aftermath of the outbreak, as countries around the world grappled with the economic fallout, the volume’s twelve contributions from leading researchers and junior scholars draw on the rich debates of the 2019 congress of the Swiss Sociological Association on the future of work held at the University of Neuchâtel in the fall of 2019. The chapters, divided into five sections, cover issues ranging from the impacts of digital technologies and globalization to the experience of marginalized workers and the future of academia. Their critical insights into the historical dy- namics and lived experiences behind the transformation of work provide a framework to understand the fate of workers and occupations in these unsettling times. Christian Suter is professor of sociology at the Universi- ty of Neuchâtel, Switzerland. His research has focused on social and economic inequalities, social indicators and quality of life, as well as on globalization and world society. Jacinto Cuvi is visiting fellow at the London School of Economics. He holds a PhD in sociology from the Univer- sity of Texas at Austin. His research focuses on informal labour in the urban South. Philip Balsiger is an assistant professor of economic sociology at the University of Neuchâtel, Switzerland. In his work he is interested in sociological approaches to markets, with a particular attention on contentious processes. Mihaela Nedelcu is titular professor at the Institute of Sociology and project leader in the nccr-on the move at the University of Neuchâtel, Switzerland. Her research ISBN: 978-3-03777-244-7 analyses the social impact of digital technologies and the transnationalization of social life, with a focus on e-diasporas, e-borders, and transnational ageing.