Master’s Thesis – Second Submission
University of Amsterdam
Research Master Urban Studies
15 August 2019
Author
KELLY DONOVAN
Supervisor
WILLEM BOTERMAN
Second reader
NIELS VAN DOORN
When customer is king and the platform is God:
place and precarity in the platform economy
Uber drivers and their experiences with insecurity in Glasgow and Amsterdam
Drawing from interviews with Uber drivers in Glasgow, Scotland and Amsterdam, the Netherlands, this study explores how local urban contexts mediate experiences with precarity. To date, nearly 4 million Uber drivers in over 750 cities operate under similar conditions of insecurity. These conditions accompany the general insecurities of independently contracted work in post-Fordist markets, as well as platform companies’ specific business practices. Drivers, more or less able to enjoy the benefits of flexible and impermanent work depending on personal circumstance, are also embedded in different urban environments whose physical, socio-economic and regulatory contexts filter the “touching down” of platform companies. Experiences with precarity amongst drivers in Glasgow and Amsterdam – two cities which sustain supply and demand for Uber differently – offer a lens of insight to labor flexibilization and risk transference in post-Fordism. On the one hand, similarities across Glasgow and Amsterdam speak to the convergence of labor experience across place. On the other hand, nuanced differences in driver experiences reveal the impact of place. Selectively focusing on the experiences of drivers in two Western European cities with different physical, socio-economic and regulatory environments, this study delivers situated knowledge on how local context intersects global restructurings, personal circumstance and platform design to mediate experiences with precarity. The socio-spatial implications of how drivers cope with their precarity, similar across Glasgow and Amsterdam, are ultimately hypothesized to be applicable to a number of urban economies.Index
Introduction………4 Theoretical framework ………..6
‘Precarity’: definition and application in the literature
Economic restructurings: flexibilization, informatization and precarity in the 21st century ‘Gig’ labor: precarity in the platform economy
Choice and constraint in ‘gig’ labor
Precarity and place in the platform economy Uber: “poster child” of the platform economy
Studies of Uber drivers to date: personal circumstance, place and precarity
Uber in Glasgow and Amsterdam: case selection and case context……….14
Glasgow introduction
Uber: ‘touching down’ in Glasgow Amsterdam introduction
Uber: ‘touching down’ in Amsterdam
Labor market structures: Glasgow and Amsterdam
Research design and methodology……….23
Conceptual model and operationalization Methodology
Data collection in Glasgow Data collection in Amsterdam Data analysis
Limitations
Empirical findings………30
Driver profiles
Learning curve: disillusionment and fear
Personal circumstance in Glasgow: capability and constraint Personal circumstance in Amsterdam: capability and constraint Influence of local context
1. Physical security in Glasgow 2. Surge pricing in Glasgow
3. Built environment and transportation infrastructure in Amsterdam 4. Extreme adaptability in Amsterdam and its racial implications Coping with precarity: implications for socio-spatial disadvantage
Conclusions and discussion……….52 Sources………...54 Appendixes………61
Introduction
From home cleaning to child care and delivery, interactive service labor in post-Fordist markets is increasingly performed by a contingent workforce. This workforce includes a growing number of laborers who perform ‘gig’ tasks disseminated online. Defined as “the exchange of labour for money between individuals or companies via digital platforms that actively facilitate matching between providers and customers, on a short-term and payment-by-task basis” (Taylor et al., 2017, p. 8), ‘gig’ labor increasingly defines 21st century work trajectories. More than 36% of U.S. workers are estimated to engage in gig labor (Gallup, 2018). This figure is expected to grow 24% by 2027 (MBO Partners, 2018). In the United Kingdom, meanwhile, 4% of the workforce (1.3 million people) takes part in the gig economy, and 12% of working-age adults indicate they have considered trying gig labor (CIPD, 2017).
Featuring low barriers to entry, flexible hours, minimal start-up costs and the ability to work ‘on-demand,’ app-based ‘gig’ labor appeals for reasons of choice and constraint. For those wishing to top-up additional sources of income, gig labor is easily combinable. In the United Kingdom, for example, 58% of gig workers report engaging in gig labor on top of permanent, more ‘traditional’ work (CIPD, 2017). Gig labor also appeals for reasons of sociability, especially among retirees (International Labour Organization, 2019). For the vulnerable and disadvantaged, meanwhile, (e.g. migrants and former convicts), gig labor represents an economic “savior” (Ravenelle, 2019). Especially in developing countries, where informal work is already common, platform companies have generated “a paradigm shift in the employment market,” leading to perceptions of gainful employment (Gururaj, 2015, p.1).
Yet concerns of “apploitation” (Calloway, 2016) challenge celebratory rhetoric of flexibility and egalitarianism. A growing body of research shows how platform companies co-opt the rhetoric of sharing and collaboration to “exacerbate the already precarious conditions of contingent workers in today’s low-income service economy” (van Doorn, 2017, p.898), making “uncertainty and insecurity…the price for extreme flexibility” (Aloisi, 2016, p.653). Although uncertainty and insecurity is not unique to the platform economy (temporary and short-term work has “long been a staple of low- and middle-income work” (Tran and Sokas, 2017, 2017, p.e63), and informal work arguably defines the majority of work in developing countries), “marginalised and casualised employment” has nonetheless “become the prevalent form of contemporary labour relations in post-Fordism” (Waite, 2008, p.416), exposing more people to “flexploitation” (Neilson and Rossiter, 2005). Additionally, flexible work in post-Fordism is notable in that “it conflates categories of workers usually at opposite ends of the labour market spectrum, low-paid workers (e.g. cleaners, drivers, construction workers, carers, domestic labourers) and the higher paid so-called ‘creative class’ (e.g. IT workers, advertising workers)” (Waite, 2008, p.417).
Platform companies, offering a myriad of short-term one-off tasks, are significant contributors to the popularization and normalization of contingent work. Of the many web-based platforms which drive digital economic circulation in the 21st century, Uber is one of the largest. Launched in 2010, the American company digitally matches people seeking taxiing services with non-professional drivers. In 2018, 3.9 million independently contracted drivers in over 63 countries were estimated to carry out 14 million rides daily (Uber Newsroom, 2019). Despite the company’s claim to being a “passive conduit” for exchange (Uber Legal, 2017), critics argue that Uber’s high risk/high reward investment portfolio is more akin to that of a venture capital fund than neutral intermediary. By not only enabling connections but also profiting from this connectivity, platforms such as Uber are argued to represent “legitimate object[s] of capitalization” (Langley and Leyshon, 2017, p.5). From $50 billion in July 2015 to $120 billion in October 2018, Uber’s valuation is particularly reflective of the extent to which platforms, “a distinct mode of socio-technical intermediary and business arrangement that is incorporated into wider processes of capitalism” (p.1), can profit from digital intermediation.
The human and social costs of this profit are relevant to a number of platforms, but Uber’s rapid growth has inspired a growing number of studies on the experiences of Uber drivers. These studies
respond to calls for more scholarship on the explicitly labor side of digital labor (Fuchs, 2014), and deliver empirical insight on the experiences of contemporary laborers under “new ways of monetizing a range of human activities” (Malin and Chandler, 2017, p.384). To date, much of this research has focused on drivers’ personal circumstances and motivations to drive for Uber (e.g. Hall and Kreuger, 2017). A growing body of research also relates uncertainty and insecurity to the platform’s design features, which limit flexibility in practice (Rosenblat and Stark, 2016) and contribute to the normalization of social interactions (Chan and Humphreys, 2017). An equally growing body of research examines how drivers cope with web-based systems of management, pointing to “strategies of arbitrage between worker autonomy and worker control” (Shapiro, 2018, p.2954). In analyzing the risks which Uber drivers bear, some have argued that “rather than ‘shifting’ risk onto workers, Uber may well be creating a new market, with a new allocation of risk and reward. How much risk drivers will bear, and what rewards they will enjoy, are very much open questions” (McKee, 2017, p.98).
Without negating the value of such research, relatively little attention has been paid to the role of
place in mediating experiences with precarity. A critical geography of precarity (Waite, 2008) is important,
however, because such a perspective could contribute to contemporary discussions of “glocalization” (Brenner, 2003), as well as advance existing theory on the spatial unfixing of work under transitions to the platform economy. As the number of platforms continues to grow, and as the variety of tasks disseminated by these platforms continues to expand, frameworks have emerged to distinguish between platforms offering virtually transmittable tasks (e.g. UpWork and CoContest) versus manual tasks requiring a local physical presence (e.g. Deliveroo and LastMinut) (de Groen and Maselli, 2016). Uber, reliant on millions of spatially-fixed drivers to carry out its core labor ‘on the ground,’ is an example of the latter. Manually interactive, passenger transportation takes place across (urban) space. The human agents who perform this labor, varying in their socio-material constraints and capabilities, are embedded in different physical, socio-economic and regulatory environments. These environments present drivers with sources of security and risk which precede the arrival of Uber, and therefore necessarily filter the ‘touching down’ of Uber in place. Drivers must assimilate these place-based variables into their daily sense-making. Exactly how drivers do so, and how these experiences compare across place, however, remains an understudied topic.
Existing place-based studies of driver experiences are mostly single cases. Surie and Koduganti (2016), for example, analyzed driver experiences in Bengaluru, India. Malin and Chandler (2017), meanwhile, focused on driver experiences in Pittsburgh, Pennsylvania. Most recently, Kashyap and Bhatia (2018), offered a second Indian case study of drivers in Delhi. Despite delivering valuable insight on how specific physical, socio-economic and regulatory environments mediate driver experiences, the tension between localized practices under global forces demands greater comparative analysis of driver experiences. Indeed, Malin and Chandler (2017) conclude with a call for comparative research, asking, “How does the experience of an Uber driver in Bombay, London, or Paris differ from that of a driver in Pittsburgh?” (p.397).
To help fill this gap, this study compares the experiences of drivers in two Western European cities. The findings deliver situated knowledge for advancing theory on the (non)spatiality of Uber, offering a comparative European complement to existing research. A comparative design allowed for tracing difference across place to better understand how experiences with precarity, adjusted to local context, are also reflective of global restructurings. As a small-n study, this research focused on two cities: Glasgow and Amsterdam. As will be shown, certain features of Glasgow and Amsterdam’s respective physical, socio-economic and regulatory environments make these cities particularly useful cases for understanding how difference across place mediates experiences with precarity.
Theoretical framework
‘Precarity’: definition and application in the literature
The use of the term ‘precarity’ in the social sciences is diverse and contested, as is its interpretation (Waite, 2008). Bourdieu (1963) is credited with first using the French version précarité to describe contingent workers in Algeria (“travailleurs intermittents”). The term did not gain immediate traction in France, where widespread unemployment – and even irregular employment – was uncommon. It was not until the latter decades of the 20th century, under neoliberal economic restructurings, that précarité was linked to employment, giving rise to the concept of the “working poor” (Waite, 2008, p.415). Precarisation – “the process whereby society as a whole becomes more precarious and is potentially
destablised” (p.415) eventually followed, after Offredi (1988) and others suggested précarité could pertain to more than just employment. The word has only recently entered English parlance, corresponding with the turn of the 21st century and economic restructurings (Standing, 2011).
Since its popularization beyond continental Europe, the term has been used to describe conditions ranging from fragility and powerlessness against oppressive leadership (Butler, 2004) to widespread fear fueled by the rhetoric of terrorist threat (Furedi 2002, 2005). Some have even argued that precarity is part of the human condition, and “is not limited to a specific context” nor “imposed by global events or macrostructures” (Ettlinger, 2007). Waite (2008) offers a basic definition of precarity as “literally referring to those who experience precariousness” and “life worlds characterized by uncertainty and insecurity” (p.416). In reviewing the term’s many applications in the social sciences, Waite (2008) also argues that “most writers on precarity see the condition and/or its potential for mobilization as occurring in uniquely contemporary times and in particular neo-liberal spaces” (p.418-419).
Economic restructurings: flexibilization, informatization and precarity in the 21st century
Under transitions to neoliberalism, flexible employment has steadily replaced ‘normal’ or ‘regular’ employment, i.e. regulated, fulltime and long-term, cutting overhead costs and shifting risk onto laborers (Zwick, 2017). An ethos of lean production and organizational decentralization sustains these changes, spurred on by the spread of the Internet. A growing body of research reviews the extent to which flexibilization and informatization have proceeded hand-in-hand to normalize labor precarity, concluding “informatization in dominant capitalist countries has fundamentally transformed the nature of jobs and employment” (Liao, 2017, p.423). The cumulative effect is “a new employment contract,” characterized by “the increased significance of external labour markets or external employees for usually only mid-term recruiting,” as well as “omni-present information technologies, total quality management, and performance-dependent ways of payment” (Schmiede, 2006, p.341).
On a structural level, the normalization of web-based surveillance and flexible work has eroded hard-won labor achievements, producing “a clear shift of power from waged work towards capital” (p.342). On an individual level, “the decreasing biographical continuity of gainful employment is not without consequences for the working people’s way of life, for the way they see themselves, and for their self-confidence” (p.341). Most notably, “long-term biographical plans are replaced by short-term or at the most mid-term perspectives…[The] increased exchangeability of workers (increase of contingent work force), weakens the individual’s position towards the company; furthermore, he/she becomes more susceptible to reprisals of open or hidden nature” (p.341).
These patterns accompany growing socio-economic polarization as a result of informatization. As work of all types moves from the material world to the digital world, high- and low-skilled laborers alike must increasingly work with data. Affinity for this data, or lack thereof, bears significant implications for employability and advancement: “altogether, when looking at the changes of work in informatized capitalism we come to the conclusion that the forms of social inequality connected to it have clearly
increased, a development which is often called ‘digital divide’ or more precisely ‘social digital divide’” (p.342). Under this divide, the erosion of the middle class is argued to be indicative of “an overlap of several tendencies of development: on the one hand, the general, average level of education and qualification is rising. At the same time, however, the fringes of the qualification spectre seems to drift apart, which results in a suction for medium qualifications. But these polarization tendencies are thirdly much more distinctive for employment conditions and job chances. Finally, they are fourthly eclipsed by a clear spread of income levels” (p.343).
‘Gig’ labor: precarity in the platform economy
Platform companies – poster children of “the new employment contract” – are important contributors to labor precarity in the 21st century. Platforms first emerged in the wake of the 2008 global financial crisis. Grounded in the rhetoric of sharing, these platforms were distinctly not-for-profit. In recent years, however, a proliferation of platforms which facilitate the selling of small jobs on a one-off basis has emerged. These platforms help fill the void created by unemployment in the face of shrinking welfare states and austerity. They also correspond with changing consumer preferences of the urban middle class (on-demand and experiential) as well as transitions to service economies.1 Today, a myriad of tasks ranging in skill and interactivity are now available for purchase at the click of a button, sold on a one-off ‘gig’ basis. Platform companies – ‘rentiers’ of the networks they create (O’Dwyer, 2015, p.234) – orchestrate this connectivity from only a few dispersed headquarters.
Referred to as ‘partners’ (Uber), ‘riders’ (Foodora) and ‘self-employed service providers’ (Helpling) – anything but ‘employee’ – gig workers are either partially or completely exempt from social protections such as sick pay and retirement benefits. Gig workers are also responsible for covering their own insurance and supplying “whatever tools or assets are necessary to accomplish their work” (Tran and Sokas, 2017). The independent contractor designation is not new to the platform economy, but platform companies could not survive without this “tool from the neoliberal playbook” (Zwick, 2017, p. 679). Research suggests that “on-demand companies would cease to be profitable were it not for the independent contractor designation: fulfilling employee obligations would simply exceed their financial capacity” (Shapiro, 2018, p.2955). The meaning of the designation is also different in the context of gig work: unlike “traditional independent contractors,” gig workers “cannot negotiate their rates or work contracts, but must electronically accept the platform’s terms in order to access assignments” (Tran and Sokas, 2017, p.e63). Finally, as growing legal challenge suggests,the classification is often misapplied.2 Nevertheless, the term’s continued application is the means by which platform companies “evade virtually every benefit attained by workers over the last 100 plus years of struggle” (Snider, 2018, p.564).
In addition to the insecurities created by the independent contractor classification designation, platform companies’ business practices also contribute to precarity. A review of gig labor in the United Kingdom, for example, referenced “reports of an oversupply of labour at certain times, effectively flooding the market and driving down the hourly rate to below that of the National Minimum Wage” (Taylor et al., 2017, p.37). Platforms and their app-based design features have also been linked to feelings of precarity. By integrating web-based tools for workforce management (e.g. secure payment systems, identity verification, geolocation and ‘peer-to-peer’ ratings), platforms limit workers’ informed decision making and maintain a close degree of surveillance. McKee (2017) likens these features to “a form of private regulation,” which negates any claims to worker flexibility and requires actors to “buy and sell according to the rules [the platforms] have designed” (p.484). In governing “all aspects of the interactions among its users,” (p.484), a high degree of subjectivity characterizes user experiences. Rogers (2015), for example, in evaluating the social costs of platforms, identifies discrimination as a risk of ‘peer-to-peer’
1 According to van den Berg and Arts (2019), the majority of workers in western welfare states now work in (interactive) service jobs. 2In 2016, a UK court ruled that Uber drivers are, in fact, not self-employed and are entitled to paid holidays and the national minimum
ratings. Although platforms did not create discrimination, their rating systems do expose employability to subjective evaluations, based on looks. In times of precarization (van den Berg and Arts, 2018), aesthetic assessment is characteristic of “a continuous labour for labour,” consisting of a constant effort to “secure a ‘fit’ between [one’s] look and work, with the purpose of remaining employable in this economy” (p.301).
Finally, experiences with precarity in the gig economy also have an emotional and psychological element; what has been called “the social pathology of flexibilized and informatized work” (Schmiede, 2006, p.341). Emotional and psychological experiences with precarity vary by profession and skill. Petriglieri et al. (2019), for example, showed that independent workers engaged in creative or otherwise knowledge-intensive experience “stark emotional tensions encompassing both the anxiety and fulfillment of working in precarious and personal conditions” (p.124). By developing connections to certain people, places, routines and sense of purpose, however, these workers are able to “manage the broad range of emotions stirred up by their precarious working lives” (p.124). The extent to which such findings are relevant to those engaged in low-skilled independent work, where quantified rating systems script interpersonal interactions (Chan and Humphreys, 2017) and limit the opportunity for personal expression, requires greater attention. Research on taxi drivers in South China, for example, shows how “precariousness negates these male workers’ sense of self” (Choi, 2018). The extent to which insecure masculinities resonate in other rapidly urbanizing areas could be relevant to Uber drivers, who are often of migrant backgrounds (Zwick, 2017).
Choice and constraint in ‘gig’ labor
Understandings of precarity as a human condition necessitate a consideration of personal circumstance and “an individual’s own agentic demands for flexibility amid a landscape of generalized precarity. Not all holders of deemed precarious jobs feel in a precarious position (Fantone, 2007) as people may experience particular life-cycle needs (part-time work combined with full-time education for example) and conversely stable jobholders may be touched by the trappings of precarity” (Waite, 2008, p.418). By the same token, the experiences of those who elect for flexible working will necessarily differ from those who lack other options. Questions of agency and structure as well as choice and constraint, thus, remain an important analytical focus within studies of precarious labor.
Platform work in particular has often been presented as consensual and non-coercive, thanks to platform companies’ invocation of liberal market ideals (McKee, 2017). In return for uncertainty and insecurity, gig workers are free to choose their own hours. Yet platforms which disseminate gig labor also emerged in a time of high unemployment and austerity following a global economic crisis. Research on the demographics of gig laborers reveals that a significant number fully depend on gig labor in lieu of other options. A 2017 survey of gig workers in the United States, for example, found that although 44% of respondents reported being employed full-time, 24% reported being employed part-time and 32% reported being unemployed (Smith, 2016). Additionally, “many work for more than one platform, patching together a living via multiple gigs” (Tran and Sokas, 2017, p.e63). And although gig workers are generally more highly educated, this type of contingent labor also appeals to those people whose personal circumstance already puts them at a disadvantage. In the United States, for example, the same 2017 survey found that “among the 30% of respondents who described the income derived from gig work as ‘essential,’ 57% had annual household incomes of less than $30,000 per year, and 52% had high school degrees or less; they were also more likely to be non-white…and to consider themselves to be employees, rather than independent contractors” (p.e64).
This data challenges platform companies’ claims to consensual transactions, suggesting instead that “people’s use of these services is not entirely free; it is a product of their circumstances and the range of options available to them. And these circumstances and options are partly constituted by the state legal system and its interactions with the platforms’ own normative arrangements” (McKee, 2017, p.491).
Precarity and place in the platform economy
Research on the technologies of globalization highlights the extent to which Internet connectivity has disrupted traditional place-based constraints (Graham et al., 2017). Under post-industrialization, the digitization of previously mechanical processes has deepened people’s connectivity to digital structures of production. The proliferation of companies which maintain online platforms for the selling of small jobs has created opportunity unbound by place. No longer is the individual worker restricted to employment opportunities in his or her local area. Equally significant, no longer is the company dependent on local labor pools. Under transitions to the platform economy, “tasks such as translations, transcriptions, lead generation, marketing, and personal assistance can now all, in theory, be done by workers from anywhere for clients based anywhere” (Graham et. al., 2017, p.137).
Yet in assessing experiences with risk and vulnerability, scholars (e.g. Bulkeley, 2001) argue for the importance of considering place-based differences. According to Waite (2008), “the term precarity has yet to make a significant appearance in human geography despite its growing profile in other areas of social science” (p.412-413). Yet the geography of precarity raises important questions, including, “can precarity ever be a common name for the conditions found in diverse spaces? Is it able to make alliances and comprehend the spatial and scalar differences that geography is so attuned to?” (p.413). By the same token, attention to precarity across place serves as a reminder that precarity is not a uniquely post-Fordist experience. Rather, “the idea of precarity is…not new at all even if it has not been specifically labeled as ‘precarity.’ It is important therefore to recognise the distinction between the increasing usage of the word and the condition itself in terms of those experiencing precarity” (p.419). This point is made especially clear by studies of labor precarity outside the global North, where “the particular development trajectories of countries in the global South have meant that the ‘precarious condition’ is rarely even noted, perhaps because it is so ubiquitous” (p.419).
Uber: “poster child” of the platform economy
The theoretical concepts outlined above will now be applied to Uber: the specific focus of this research and a “poster child” for the platform economy (Zwick, 2017). As a “transportation network company,” Uber’s growth is illustrative of a contemporary trend of ‘splintering urbanism’ (Graham and Marvin, 2011). Under shrinking welfare states, reduced public funding and “the outsourcing of various infrastructures from public to private management” (Malin and Chandler, 2017, p.387) has created openings for private companies such as Uber to “fill in gaps created when buses cannot adequately serve a city’s transportation needs” (p.387). The implications for labor precarity are significant, because “in defunding public transportation, municipalities have essentially outsourced their transportation infrastructures to Uber…who then outsource that infrastructure again to what is essentially a group of consultants” (p.387).
Together, these trends are reflected in the ‘touching down’ of Uber in place. Despite operating in over 750 cities, Uber does not operate everywhere. Where Uber ‘lands’ is neither random nor coincidental. Certain physical, socio-economic environments and regulatory environments are more or less suited to the introduction and integration of taxi aggregators such as Uber. Indeed, the fact that Uber has been banned from certain cities and countries where it once operated illustrates the effect of regulatory contexts on filtering the ‘touching down’ of Uber in place. Yet on the whole, Uber’s remarkable geographic spread reflects the appropriation as well as acceleration of global restructurings whose impact on changing patterns of production and consumption are mirrored in a growing number of societies (e.g. more experiential and on-demand consumerism, as well as the normalization of contingent work).
With its low barriers to entry and no set working hours, the Uber platform is relatively accessible and flexible. The company does not require language fluency nor proven geographical knowledge. It also
gives drivers the option to work up to 10 hours/day, and advertised wages are considerably above minimum wage. The technology also sounds simple enough to use: “once online, you’ll automatically begin to receive requests in your area. Your phone will sound. Swipe to accept” (Uber Drive, 2019c). Especially for those whose employment opportunities are limited to begin, driving for Uber presents a relatively attractive alternative to unemployment or other precarious arrangements.
In an industry where taxi drivers face up to 15 times the average exposure to occupational violence (Mayhew, 2000),3 Uber’s structural design features also encourage feelings of safety. The Uber service is “designed to be an entirely cashless experience in most cities” (Uber Help, 2019). Lack of cash reduces the risk of robbery and requires riders to have a minimum level of financial stability (McKee, 2017). The rating system, meanwhile, sets the impression that problematic customers are deactivated. Drivers and customers rate each other 1-5 stars based on ‘Community Guidelines,’ intended to make the Uber experience “enjoyable and safe for everyone” (Uber Legal, 2019). In effect, using the platform requires riders and drivers alike to consent to the guidelines, even if one would not necessarily abide outside of the platform. By establishing its own rules of engagement to “ensure that everyone is accountable for their behavior” (Uber Legal, 2019) and deactivating users in case of disobedience, the Uber company acts as a digital mediator of virtual and physical space (Chan and Humphreys, 2017).
Yet any benefits or securities which the platform purports to offer comes at the expense of significant risk. As independent contractors, drivers are partially or completely exempt from social protections. Drivers also lack collective representation. The company “reserves the right to unilaterally alter the terms and conditions of the contracts under which its drivers work” (McKee, 2017, p.49), changing its prices and commission at will (Huet, 2014b). Drivers, lacking bargaining power, have little ground for protest. While these risks are not unique to Uber drivers, they challenge traditional understandings of self-employment, which grant some negotiating power (Tran and Sokas, 2017). Socioeconomic insecurity is further compounded by variable earnings combined with high expenditures. Compared to other types of gig labor, e.g. home cleaning or child care, “the privilege of using the driver app” (Malin and Chandler, 2017, p.385) comes at a high price. Drivers must pay city licensing fees as well as insurance, vehicle maintenance, fuel and other operating costs. These costs should be manageable under promises of hourly earnings above minimum wage, but 25% commission per ride combined with over-supply and a strategy of “‘two-sided markets,’ whereby intermediaries extract surplus from one side of the market (in this case, drivers) in order to subsidise participation on the other side (passengers)” (McKee, 2017, p.488), explain why in reality, most drivers make less than minimum wage (Huet, 2014a).
Furthermore, flexibility is in name only. In practice, the platform’s design features restrict drivers’ ability to enjoy the benefits of self-employment in practice. For example, the platform requires drivers to use its integrated navigation system, storing data on every ride. Promising workers flexibility and autonomy, yet collecting data on its drivers’ every movement, Uber has an impressive degree of scope over how drivers move across time and space. The algorithmic ‘nudging’ system, meanwhile, encourages drivers to work at specific times and travel to areas where demand exceeds supply, incentivizing drivers with the lure of higher demand and ‘surge’ pricing. Whereas nudging in other contexts simply represents informational advertising, predictive notifications of demand challenge the company’s claim that its pricing mechanism reflects “demand in real time” (Uber Drive, 2019b). Furthermore, structural information asymmetries impair drivers’ ability to judge the reliability of these recommendations. Lacking “key pieces of information that would otherwise help them make informed choices about their decisions” (Rosenblat and Stark, 2016, p.3771), drivers do not know whether nudges are high- or low-confidence recommendations.
3Victimization in the taxi industry ranges from vandalism to robbery, fare evasion, false allegations, hijacking and assault (Stenning,
1996), with most physical injuries consistent with attacks from passengers in the rear seat (Haines, 1997). Risk factors are standard to the industry, regardless of place. They include working alone, dealing with strangers in close range and carrying cash (Mayhew, 2000), however risk is typically higher in inner-city and low-income areas (Stone and Bienvenu, 1995). Female drivers are particularly susceptible to sexual assault (Alexander, Franklin and Wolf, 1994), and ethnicity can also heighten risk (Adams, 1996).
Threat of deactivation further compounds precarity. The company assesses quality of service based on cancellation rates, acceptance rates and customer ratings. “High-quality drivers typically have a cancellation rate lower than 5%” (Uber Legal, 2019), and drivers can be deactivated if their cancellation rate exceeds their respective city’s maximum rate (the average cancellation rate of all drivers in that city). The direct link between peer-to-peer evaluations, biased in any number of ways, and employability, however, is a significant source of risk, especially when temporary deactivation can escalate to permanent deactivation: “if your cancellation rate continues to exceed the maximum limit, you may lose access to your account” (italics added) (Uber Legal, 2019). Low acceptance can also lead to deactivation: “if you consistently decline trip requests, we will assume you do not want to accept more trips and you may be logged out of the app” (italics added) (Uber Legal, 2019).
Lack of clarity concerning when the company will or will not take disciplinary action and the permanence of such action hangs as a threat over drivers. This threat is compounded by the limits which the platform places on drivers’ ability to make informed choices regarding which requests to accept or cancel. For example, drivers can deny requests, but the platform operates a closed-loop system and reveals the destination only after drivers pick up the customer. Intended to prevent drivers from discriminating against short trips or stigmatized areas, the closed-loop system has exposed drivers to trips upwards of 9 hours for which they might earn less than 45% of the total fare (Griswold, 2017).
The platform’s rating system is an especially large source of insecurity, because it demands extreme adaptability. Under transitions to service economies, markets have witnessed “fundamental changes in what is required of workers when compared to unionized industrial jobs” (van den Berg, 2019, p.3). Today, service encounters are built around a performative element, whereby the goal is to “look, sound and move in a certain way” (p.2). Within ride-hailing, taxi drivers have long solicited tips based on their ability to manage affective relationships (Davis, 1959). Under informatization and organizational flexibilization, however, a growing number of companies has introduced digital rating systems to shift monitoring duties outside company walls (Schor and Attwood-Charles, 2017). Now, “Uber possesses much more information about service encounters than drivers and can use the information to script drivers’ performance through the deployment of data-driven technologies” (Chan and Humphreys, 2018, p.33). Not only tips but also basic earning potential depends on drivers’ ability to engage affectively. In the context of Uber, access to the platform is “determined in part by [one’s] ability to smile and appear cheerful” (Rogers, 2015, p.101). Drivers are expected to read their passengers’ mood and adjust their own behavior accordingly, sometimes in less than a few minutes. The extent to which drivers have successfully performed their affective duties are reflected in customer ratings. Supposedly ‘peer-to-peer’ and therefore egalitarian, this rating system enables information and power asymmetries.
Uber’s Community Guidelines call the ratings a “two-way system,” but riders can choose to rate while drivers must rate every rider. Riders are also not deactivated if their average rating falls below a certain number. For drivers, however, dropping below a certain minimum is grounds for deactivation.4 The guidelines do not specify at which point deactivation becomes permanent: “if your average rating still falls below the minimum after multiple notifications, you will lose access to your account...We may allow you to regain access to your account if you can provide proof that you completed one of these quality improvement courses” (italics added) (Uber Legal, 2019). Ratings are calculated according to the last 500 rated trips (or total number of rated trips, if under 500). Considering that only 65-70% of riders leave ratings (RideGuru, 2018), resetting one’s rating can require many more than 500 rides. Under this system, each rating essentially “situates drivers’ performance relationally: a driver’s rating reflects not only one’s performance but also the extent to which one’s performance conforms to the standard norms of ‘ideal’ performance held by other drivers” (Chan and Humphreys, 2017, p.33).
The room for subjective assessment further compounds precarity. As fitting with research on aesthetic evaluation practices in post-Fordist labor markets (e.g. van den Berg, 2019), drivers are told they can improve their ratings by maintaining a clean car and dressing professionally (Uber Help, 2017). Suggestions for sociability are also inherently subjective, e.g. keeping conversations “polite, professional and respectful” (Uber Help, 2017). Deactivation assumes drivers “no longer meet the overall quality standards that riders reasonably expect from drivers when using the Uber app in the relevant city” (italics added) (Uber Legal, 2019). Assessing reasonability, however, is deeply personal and cultural (Rosenblat et al, 2017). What might be unreasonable for one person could be perfectly acceptable for another.
Access to work based on aesthetic is not unique to Uber. A growing body of research has examined how embodied characteristics such as tattoos (Timming, 2015), accents (Nath, 2011) and even smiles (Hochschild, 1983) influence employability. In interactive service labor, securing the right ‘look’ has been called “part of a continuous labour for labour” which feeds into “a performance of flexibility, employability and a display of a desire for paid labour” (van den Berg and Arts, 2019, p.301). Under transitions to the platform economy, the bar has been raised even higher. Now, assessments are conducted not by a qualified hiring manager, but by strangers whose contact with service laborers might consist solely of virtual communication. In the case of Uber, drivers are continuously assessed and reassessed by strangers whose contact with drivers might only last two minutes and consist of a few words. The implications are especially significant for minority drivers. In today’s urban economies, where ethnic minorities increasingly perform manual service labor, driving for Uber requires more than just delivering passengers safely. To avoid deactivation, minority drivers “may need to overcome white passengers’ preconceptions, which can involve ‘identity work,’ or a conscious effort to track white, middle-class norms” (Rogers, 2015, p.97-98).
The ways in which drivers perceive constraints to their autonomy and informed decision making in practice adds to the growing body of research on how high- and low-skilled workers alike deal with ‘not knowing’ in the age of information. Although the platform’s design features have been shown to “premediate expectations of service encounters and spatial movement” (Chan and Humphreys, 2017, p.29), leading drivers to adapt their practices in ways which conform to the company’s logics, research also suggests that contemporary systems of control “ultimately hinge on workers’ willingness to conform to the calculative rationalities that companies project onto them” (Shapiro, 2018, p.2954). Rather than responding to incentives such as surge pricing based on pure wage maximization, research suggests that Uber drivers do not act solely based on wage maximization. Determining the most salient variables in drivers’ decision making remains an underexplored area of research.
Studies of Uber drivers to date: personal circumstance, place and precarity
In 2017, Hall and Krueger produced a comprehensive analysis of Uber drivers in the United States. To overcome the difficulties of extreme variety in the number of hours which Uber drivers work, the study considered ‘active partners’ to be those who provided at least four rides per month. Based on this classification, Hall and Krueger found the following:
A tremendous amount of sorting takes place in the on-demand economy, and, by dint of their backgrounds, family circumstances, and other pursuits, Uber’s driver-partners are well matched to the type of work they are doing. Notably, Uber’s driver partners are attracted to the flexible schedules that driving on the Uber platform affords. The hours that driver-partners spend using the Uber platform can, and do, vary considerably from day to day and week to week, depending on workers’ desires in light of market conditions. (p.706)
Additional studies present more nuanced driver profiles, adjusted to personal circumstance and local market conditions. Existing case studies on taxi aggregators such as Uber and Ola show that “a market for Uber and Ola already existed” and that these companies “do not aim to merely appropriate
the existing market but expand it, by creating additional demands” (Kashyap and Bhatia, 2018, p.175). In Pittsburgh, Pennsylvania, for example, liberal restructurings account for “a highly educated but underpaid workforce that seems especially suited to the ‘flexible labor markets’ (Benner, 2002) of the high-tech economy” (Malin and Chandler, 2017, p.383). Malin and Chandler also found that Pittsburgh’s drivers are “relatively privileged in relationship to Pittsburgh’s mostly African American jitney drivers who have provided important ‘illegal’ taxi services at least since the 1970s” (p.386). Compared to the jitney drivers, Uber drivers “enjoy a range of institutionalized privileges and protections” (p.386). Even so, precarity is conditioned by personal circumstance and place-based variables. For example, most of Chandler and Malin’s interviewees drive for Uber as a supplement to other jobs or activities (e.g. studying). This bias explains why Chandler and Malin’s interviewees “overwhelmingly see their work as something flexible, fun and even beneficial to the larger society” (p.384), and accept Uber’s displacement of risk onto drivers as “a necessary component of Uber’s economic strategies” (p.386). Gendered differences, however, emerge in the context of navigating “the late night streets of Pittsburgh’s bar scene” (p.384), when surge pricing is common and earnings are likely to be higher. Drivers report different strategies for handling drunken aggression, depending on gender. Yet the company’s failure to respond to reports of such aggression is ultimately indicative of “the company’s continuing strategy of risk transference” (p.395).
In India, meanwhile, where “the traditional Indian cultural ethos of sharing is viewed as a congenial context for the growth of the emerging sharing platforms that will address the persisting problem of inadequate infrastructure” (Kashyap and Bhatia, 2018, p.170), Uber has found its largest market outside the United States (p.175). Despite the fact that nothing is shared on the Uber platform (Rogers, 2015, p.87), and Uber does not measure up to its expected potential in contributing to ecological sustainability (Kashyap and Bhatia, 2018), demand is nonetheless sustained by India’s overburdened public transportation systems as well as the rise of “the middle class in post-liberalization India as a class that has disposable income and indulges in a variety of forms of consumption such as education, health, shopping, cinema, eating out, travel and luxurious modes of transport” (p.181). In Delhi, Uber answers to “an insatiable demand for transport that the DMRC cannot on its own meet” (p.172-173),5 and also creates viable employment opportunities. Kashyap and Bhatia (2018) propose five profiles of drivers, differentiated by personal circumstance. For the chronically impoverished and stigmatized, for example, “it is a matter of relief...that when they are onboarded, their religion, and caste is never asked in documents as a policy” (p.177). For those who previously worked in ride-hailing, meanwhile, “shift[ing] to Uber…is a way up the socioeconomic ladder” (p.177). The educated and highly-qualified, meanwhile, e.g. salespersons who quit their jobs and joined Uber, represent “the why not drive a taxi category” (p.177).
In Bengaluru, meanwhile, Uber has found an easier point of entry than in Indian cities with well-established taxi unions, because “Bengaluru has always had a large fleet of private taxis without influential and cohesive transport associations” (Surie and Koduganti, 2016, p.7). This regulatory environment explains why “platform economy services faced no resistance when they entered into Bengaluru when compared to New Delhi and Mumbai with unions” (p.7). Surie and Koduganti (2016) apply a conceptual tool of temporality to analyzing drivers’ perceptions of security and risk, concluding that Uber has “given drivers a stable, mid-term period of time to accumulate wealth, which in turn has allowed them to stabilize and take short-term decisions by making large investments in their work, and to bear the risks of flexible working conditions in the short-term with more confidence” (p.1). Surie and Koduganti (2016) qualify the generalizability of their findings, however, because of place-based differences in India’s developmental history: “while platform economy companies are global and follow similar labor practices across their global markets, individuals on the platform in India cannot be easily compared to their global counterparts because of the flexible work options workers in India have always had” (p.3).
Uber in Glasgow and Amsterdam: case selection and case context
Glasgow and Amsterdam are especially useful locations for comparing Uber drivers and their experiences with precarity. Embedded in post-industrial economies and shrinking but nonetheless generous welfare states, Glasgow and Amsterdam are similarly sized and feature liberalized but regulated taxi markets. Both cities are also embedded in national contexts which are European outliers to flexible employment. In addition, Uber follows similar economic and labor practices in both cities (e.g. electronic payment only, and a .2 difference in the average rating which drivers must maintain to stay active on the platform). Yet Glasgow and Amsterdam are also in markedly different points in their de-industrialization process, accounting for a different spread of socioeconomic deprivation as well as different positionings in global markets. Different planning legacies and regulatory contexts, meanwhile, account for different transportation infrastructures and mobility cultures. Together, these elements provide the contrast needed to understand how drivers’ experiences with precarity are adjusted to place.
Glasgow introduction
Glasgow is a compact city which lies at the heart of the Strathclyde metropolitan area. Home to about 600,000 people, nearly 22% of which are students, Glasgow is Scotland’s most densely populated city. Very little remains of the city’s medieval street plan. A grid plan north of the River Clyde – introduced in the early 1800s – characterizes Glasgow today. Central Glasgow is home to the city’s main cultural venues and higher education institutions. (The University of Glasgow lies outside the center in the West End, a leafy area with a bohemian flair). The main shopping areas are located to the south and west, and the commercial and residential districts are located in Merchant City to the east. Glasgow has a “mature network of public transportation” (Cooper, 2007, p.199), including bus, subway and rail services, but transportation service drops significantly in the evening hours. The city’s subway, for example, which only covers 10.5km, stops at 6pm on Sundays (11pm during weeknights).
Figure 1: Map of Glasgow City Center
Once a powerhouse in manufacturing and the “Second City of the Empire” (Robbins, 1990), Glasgow has suffered from acute post-industrial decline. Nearly 50% of Glaswegians live in Scotland’s most deprived areas (Macintyre et al., 2008). The city’s highly fragmented spread of deprivation (Figure 2) dates back to planning decisions taken in the mid-20th century, namely “skimming the cream”; “rehousing skilled workers in new towns, and leaving the poorest behind” (Goodwin, 2016). Isolated and residential, these areas bred poverty and crime. As such, Glasgow quickly earned a reputation as “Murder Capital of Europe” (Kenyon, 2018). The city’s reputation for a high incidence of violent crime has only recently started to lessen. Yet compared to other UK cities which deindustrialized around the same time (e.g. Manchester and Liverpool), Glasgow still features lower life expectancy and poor health; a phenomenon referred to as “the Glasgow effect” (Goodwin, 2016).
Figure 2: Glasgow City data zones by Scottish index of multiple deprivation, 2006
Source: Macintyre et al, 2008
Ride-hailing in Glasgow has a well-established history. Most locals in Glasgow are accustomed to ride-hailing as a means of mobility, thanks in part to the city’s car-centric design. In the 1960s, large sections of Glasgow were destroyed to make way for a new motorway, the M8. This dramatic example of modernist planning came at the cost of historic architecture and long-established neighborhoods, leaving what some call “a deep scar on the nation’s largest city which has never fully healed” (McClean, 2017). For better or for worse, the M8 motorway normalized automobile transport in Glasgow, including ride-hailing, to an extent which is not spread evenly across Scotland. One Glaswegian Uber driver, for example, when asked if he uses Uber as a customer, responded, “No, I don’t drink, so I never need a taxi. But
it’s the culture of things as well. I was brought up in Edinburgh. And taking a taxi there, it’s seen as a wee bit extravagant. Whereas in Glasgow it’s definitely part of the culture. No question about it.” According to the Chair of
Glasgow City Council’s Licensing and Regulation Committee, Glasgow was well-suited to support the arrival of Uber. The company’s arrival has rewritten the vocabulary of mobility: “We no longer have, ‘picking
up a cab’ or whatever. It’s always, ‘I’m going to grab an Uber.’ And that shows you how much Uber has grown and it shows you how big they’ve got and how much they’re used. It’s no longer a case of, ‘I’m going to grab a taxi.’ I learned ‘I’m going to grab a taxi,’ from my very young days, and then it was, ‘I’m going to grab a private hire.’ Now it’s become ‘grab an Uber.’” Although Glasgow has only recently started to invest in diversifying modes of transport, car and
taxi still dominates. 30% of students, for example, are driven or travel by taxi to school, compared to 23.9% nationally (Glasgow Centre for Population Health, 2017).
Glasgow is home to various ride-hailing vehicles. The Scottish government distinguishes between taxi and private hire, classifying taxis as black hackney cabs which pick up from the street and private hire as pre-booked services (including Uber). As of 2018, taxi licenses and private hire licenses in Glasgow totaled 2,482 and 5,042, respectively (Scottish Government, 2019b). Licensed drivers can use the platform at will, making it impossible to know the exact number of Uber drivers at any given time. Type of ride-hailing vehicles differ in status, reputation and barriers to entry. Licenses for black cabs are restricted, and can cost upwards of thousands of pounds. Black cab drivers are also the only drivers who can pick up from the street and are the only drivers who must pass a geographic test of the city to receive their license. While not particularly renowned for their friendliness, thus, black cab drivers nonetheless possess years of accumulated knowledge of city landmarks and routes.
The liberalization of taxi transportation in the UK started with the 1985 Transport Act. Historically, taxi provision in the UK has followed a different set of rules than bus and train provision. Rather than direct supplier, the UK government has historically acted as a regulator of private taxi providers. Even so, the ’85 act moved to partially deregulate restrictions across three domains (quantity, quality and economic) and three strata (ranking, cruising and pre-booked). But the uptake has been slow. Across the UK today, Glasgow included, the majority of metropolitan areas uphold restrictions in the form of price controls, quantity restrictions and entry requirements based on vehicle age, appearance and safety (the latter still being upheld in all metropolitan areas) (Cooper, 2007). The result is a liberalized, but regulated, taxi market, which provides openings for private hire. The introduction of private hire challenged the black cab trade’s monopolistic hold on taxiing, introducing lower fares and the possibility to pre-book. Yet private hire has a largely negative reputation, with problems ranging from pirating (taking customers off the street and setting their own prices), lack of reliability (being late or sometimes never showing up at all) and corruption (bribes to operators being commonplace). In light of these concerns, the “stated aims of the city are to encourage a continuation of a high quality taxi service in parallel to identifying issues that affect the operational effectiveness of taxi services in the city” (Cooper, 2007, p.228).
Uber: ‘touching down’ in Glasgow
In November 2015, Uber arrived in Glasgow, offering an cheaper and more reliable alternative to other private hire companies. In the eyes of the council, Uber delivers an important complement to existing transportation services. Glasgow’s current infrastructure is outdated and insufficient: “We do see
there is a need for all forms of transport through Glasgow. And it would be good for Glasgow to have a better infrastructure in regards to the underground system…our transport system I would say needs upgrading with regards to subway and the bus fleets etc. to be more accommodating and to have newer vehicles” (Chair of the Licensing and Regulatory Committee,
Glasgow City Council). Until this upgrading happens, Uber helps fill a gap in service: “when you’ve got
outlying areas where you’ve not got such great bus infrastructure then you do rely on the taxi and private hire trade” (Chair
of the Licensing and Regulatory Committee, Glasgow City Council).
Additionally, Uber serves an important function in Glasgow’s nighttime economy. Near the end of the 20th century, the “twenty-four hour city concept” grew in popularity. Intended to revitalize and create safer city centers after 5pm, the concept’s UK origins can be traced to Comedia’s 1991 Out of Hours report and Manchester’s 1993 More Hours in the Day initiative (Heath, 1997). The concept has
since spread to most UK cities, Glasgow included, and “is influenced both by those cities in continental Europe which are inherently twenty-four hour in their nature and those which, since the 1970s, have developed cultural policies to revitalize their urban night-life (Bianchini, 1995). Indeed, an explicit aim of many British cities is to become a ‘European’ city” (Heath, 1997). In Glasgow, the concept’s introduction was motivated primarily by reasons of safety and city image (Heath, 1997).
Safety and city image are sensitive themes in Glasgow’s urban (re)development which also affect ride-hailing. Cooper (2007), in reviewing the taxi markets of 52 mid-sized UK cities, found that Glasgow is “seeking to establish infrastructure in which nighttime activities can be enhanced, protected, and carried out with a minimum of conflict between individuals, residents and visitors” (p.107). Cooper (2007) also found “a strong link between nighttime activities and the use of taxis as a method of returning home, taxis being the predominant mode of transport returning from nighttime entertainment” (p.131). Yet insufficient taxi supply at night, despite peak in demand, is attributed to antisocial behavior associated with the nighttime economy. Fear of passenger behavior, costs associated with cleaning and repairing damaged property as well as avoidance of payment contributes to insufficient taxi supply at night, leading to “delay in engaging taxis in excess of half an hour, and in some instances in excess of 45 minutes. Conversely, for much of the daytime a significant surplus of taxis exists” (p.132).
Within this context, the introduction of Uber in Glasgow has been well-received for reasons of supply and demand. In terms of supply, the platform’s low barriers of entry appeal to the low qualified and economically disadvantaged. This includes the nearly 1/5 of working-age residents in Glasgow who lack formal qualifications (Lalani et al, 2014), as well as the 53% of unemployed ethnic minorities (Glasgow Centre for Population Health, 2016). Taxiing in industrialized societies has long appealed to the economically vulnerable (Piore, 1979), Glasgow included:
[Some] people, they see it as an easy job to get, because there’s nothing, you don’t have to do any written exams or anything like that. It’s not as if you’re having to go through any interview process as such that you would for a normal job. You don’t have any competency-based papers to complete, so as long as you can drive and hold a license at the time then you can apply.
- Chair of Licensing and Regulatory Committee, Glasgow City Council In a city where ride-hailing is normalized, Uber has earned a reputation for even lower barriers to entry:
I don’t know what kind of background checks Uber do, and people do see an Uber license as easier to get than another license...Uber seems to be more, “we’ll take you on, we’ll give you a chance.”
- Chair of Licensing and Regulatory Committee, Glasgow City Council Besides offering the economically vulnerable a relatively accessible alternative to unemployment or otherwise precarious employment, Uber also reduces the risk of occupational violence. In Scotland, 42% of violent crime involves an offender under the influence of alcohol, and 34% occurs during peak taxiing hours: weekends between midnight and 6am (Scottish Government, 2018c). Unsurprisingly, Scottish taxi firms report disproportionately high rates of theft, threat, fraud and attack (Burrows et al., 1999). For those who have previously worked in ride-hailing as well as those who are new to the industry, narratives of enhanced bodily security by driving for Uber represents a significant motivation. Uber’s cashless payment offer some protection against theft. The rating system, meanwhile, gives the impression that problematic customers are eventually deactivated. Private hire, consequently, being cheaper than black cabs but still operating with cash, is reputed to attract the city’s more deprived population who lack access to credit. Drivers described those who do not use Uber as “head bangers” (i.e. easily violent) and/or likely to carry out illicit activities (e.g. drug transfers), enabled by the fact that black cab and private hire
do not keep a record of transactions in the way Uber’s mobile phone application does.6 Uber’s arrival has given drivers the confidence to work shifts they otherwise might have avoided: 90% of private hire drivers now work on weekend evenings, compared to 60% during the week (Hunter, 2019).
Together, these factors help explain why Uber has found such a receptive labor supply in Glasgow. It is no coincidence that the number of private hire licenses increased 31% after Uber’s arrival, compared to an average growth rate of 1.6% over the four years prior (Transport Scotland, 2019). In terms of demand, using Uber with a group is often cheaper than public transportation, which attracts students. For the average Glaswegian, meanwhile, used to queuing for taxis or waiting for a pre-booked ride to ultimately not show, Uber represents a cheaper and more reliable mode of transport. Finally, for residents of peripheral housing estates, where greater deprivation correlates with lower car ownership (Pacione, 2004) and more limited public transportation access (Macintyre et al, 2008), Uber improves connectivity.
Amsterdam introduction
Like Glasgow, Amsterdam is relatively small and dense. Home to about 850,000 people, Amsterdam is the Netherland’s most densely populated city and lies at the heart of the Amsterdam Metropolitan Region (AMR). Also like Glasgow, Amsterdam is a popular student city. Unlike Glasgow, however, the center of Amsterdam has retained much of its medieval street plan, accounting for a maze-like mix of streets, tram rails, bike lanes and bus lanes surrounded by an A10 ring road.
Figure 3: Amsterdam City Center
Source: Maps of Europe, 2019
6 When Uber launched, it was the first to introduce the practice of rating customers and drivers via an online mobile application. This market distinction is beginning to change, as private hire companies develop their own mobile application system similar to Uber’s.
Also unlike Glasgow, Amsterdam has more successfully transitioned to a service economy following deindustrialization (Kloosterman, 2015). Traditionally welcoming of foreign trade and investment, the Amsterdam Metropolitan Area is a key node in global economic networks and continues to attract international companies in the financial, creative and knowledge-intensive industries (GAWC, 2012; Musterd et al., 2006). As high-skilled knowledge workers have flocked to Amsterdam, often settling in central neighborhoods (Sleutjes and Boterman, 2016), low-skilled laborers have followed in suit. Increasingly from a nWestern background, these low-skilled laborers cater to the material and on-demand needs of their professional and managerial counterparts. As such, job growth across Amsterdam is skewed towards high and low-skill service positions, the latter distributed mostly by temporary work agencies (uitzendbureaus) (Gemeente Amsterdam Economische Zaken, 2018).
Growing concern over polarization – reflected economically, via a widening income gap; ethnically, via diverging job prospects between native Dutch and non-Western migrants; and spatially, as the A10 road increasingly separates “the rapidly gentrifying inner city neighborhoods from the relatively downgrading garden cities at the municipality’s peripheral boundary” (Savini et al., 2016, p.107) – is part and parcel of Amsterdam’s transition from “just city” (Uitermark, 2009) to “global city” (Sassen, 1991; Kloosterman, 2015). A traditionally strong welfare state and complementary safety net at the municipal level have tempered the social and economic inequalities normally associated with transitions to “a knowledge-intensive, flexible, and internationally-oriented economy” (Kloosterman, 2015, p,126). Yet cracks are beginning to show as Amsterdam has witnessed “the rise of local pockets of cultures of poverty outside of the mainstream society” (p.127).
Growing inequality is even evidenced by mobility data. In a city where 58% of residents traverse the city’s 767 km of bike lanes daily (I Amsterdam, 2019), cyclists have special priority in mobility plans, even though the normalization of biking is not spread equally across the city. Biking is less common in peripheral post-war neighborhoods, for reasons of practicality (there is more space for cars, and scooters/mopeds shorten commutes) and demographics (non-western migrants tend to live in peripheral neighborhoods, and cycling is not a cultural norm for these populations) (Figure 4).
Figure 4: Mode of transport weekday transport per city district, Amsterdam 2015
Uber: ‘touching down’ in Amsterdam
Featuring an entrepreneurial business culture and a commitment to smart city development (Peck, 2012), Amsterdam was well-primed to be receptive to Uber’s rhetoric of sharing and environmentally-friendly mobility. Given the city’s thriving business services sector, it is also no coincidence that Amsterdam houses one of Uber’s four international bases alongside the headquarters of other FinTech giants. In terms of labor supply, Amsterdam is embedded in a national context which is a European outlier to flexible employment. Temporary employment and self-employment in the Netherlands exceed European averages (OECD, 2019). Growth in self-employment consists mostly of
zelfstandigen zonder personeel (self-employed without personnel), which includes Uber drivers. Only 5% of
Amsterdam’s ZZPers are engaged in transport and logistics (Gemeente Amsterdam Economische Zaken, 2018), but recent growth in these sectors has consisted largely of new ZZP’ers who perform taxiing and delivery services (Gemeente Amsterdam, 2018).
Yet in terms of demand, car transport and ride-hailing in Amsterdam is less normalized than in Glasgow. Ride-hailing is expensive and seen as a luxury (European Commission, 2016b). Given the choice, most locals will bike, “come rain or shine” (Gemeente Amsterdam, 2018, p.3). Biking is particularly well-suited to Amsterdam’s flat topography (a feature of the Netherlands in general) as well as the city’s maze-like streetscape; a navigational nightmare for cars. As such, Uber in Amsterdam appeals to an ‘ultra-niche’ clientele, consisting mostly of tourists, expats and those people less embedded in the biking culture. Finally, given the high cost of taxis in Amsterdam (Table 1), Uber offers a cheaper alternative (Appendix 4).
Table 1: Cost of a standard 5km taxi ride in European cities
Source: European Commission, 2016a
City Cost (€) Copenhagen 19.9 London 17.4 Amsterdam 15.6 Brussels 13.9 Berlin 12.3 Dublin 11.0 Rome 10.7 Madrid 9.3 Barcelona 9.2
Uber has operated legally in Amsterdam since September 2015. The Netherlands, Amsterdam included, has a long history of taxi regulation and deregulation. Prior to 2000, the Netherlands upheld a fixed number of taxi licenses. In 2000, however, a new law (WP2000) opened the market to more competition, replacing license caps with new quality controls (Pelzer et al., 2019). In 2012, the municipality of Amsterdam introduced its own Taxi Regulation based on WP2000. This Taxi Regulation added extra quality controls (Table 4) in an attempt to correct for the fact that Amsterdam ranked in the bottom 3 of 22 European cities in a 2011 inquiry of taxi quality (ANP, 2011).7
It was against this backdrop that Uber introduced its UberBlack service in Amsterdam in 2012. At the time, UberBlack did not fully adhere to regulations (e.g. customers were not given printed
7 Complaints include drivers missing red lights, taking excessively long routes, refusing passengers for short journeys and illegally adjusting tariffs.
