key: cord-0766392-5iwo6tlb
authors: Vieira, Tiago
title: The lose-lose dilemmas of Barcelona’s platform delivery workers in the age of COVID-19
date: 2020-12-31
journal: Social Sciences & Humanities Open
DOI: 10.1016/j.ssaho.2020.100059
sha: 5ca38ac5604d85adbd528287a98f3e2881eecb3b
doc_id: 766392
cord_uid: 5iwo6tlb

The abrupt lockdown experienced by a big part of the world population due to the COVID-19 pandemic has bestowed upon home delivery services an unexpected importance. Officially considered amongst “essential services”, their workers circulate freely while most people are advised (when not forced) to stay in their homes. The present paper explores how this context helps to shed light on the precarious situation of the majority of the platform delivery workers (PDW). This is done through in-depth, semi-structured interviews and digital ethnography of the interactions within a WhatsApp group. The main finding is that the COVID-19 context deepened the precarization of the PDW confronting them with four dilemmas from which there is no way out.

The uncontrolled spread of the COVID-19 in Spain led the Government to enforce a nation-wide lockdown on March 15. Amidst critics from Autonomous Communities' leaders, opposition parties and even members of the Government itself, the enforcement of the lockdown dictated a staggering breakdown of the everyday routines of the Spanish people(s).

As has been the case almost everywhere else, the main aim of the lockdown was to prevent the proliferation of the virus, through the refraining of human contact. In order to achieve such a goal, all activities considered less than essential for facing this unprecedentedly critical moment were first advised and, later on, forced to come to a stop.

On top of this, the very ways in which people interacted became the subject of scrutiny from the authorities: any form of not strictly functional social engagement was deemed unwelcomed -even the once overcrowded supermarkets were now vast empty corridors, as people queued by the door in long, 1-m-of-distance-between-each-person-please lines.

As millions of other people in the city of Barcelona followed the protocols, I also did my best to follow the rules, only leaving my thirty square meter flat every other day, and only to buy bread and some groceries. Thirty minutes of freedom and, if lucky, some sunlight. While I briefly strolled the never-to-be-seen-so-empty-streets of the up-untilrecently bustling streets of Raval, the only few other freedom companions I'd get would be: police officers (dozens of!), dog walkers, fellow groceries' purchasers, and bike riders floating around with huge, fully colored backpacks on their backs.

All the three former categories were expected, they all had a reason to be where they were. However, the latter posed me an intriguing puzzle: in a moment where the majority of the population was notably (and perhaps justifiably) afraid of contagious contacts with other human beings, why and under which circumstances (incentives, if you will) would these workers, who are formally free to choose the extent of their availability to work, expose themselves to the possibility of hitting the COVID-19 infected list?

In the present paper, I explore the situation of the platform delivery workers (hereafter, PDW) in the peak of the COVID-19 crisis in Spain. The PDW form part of broader group of workers whose activities are performed directly and/or mediated by a digital platform, commonly known as platform workers. In the case of the PDW, the platform ensures the liaison between the users/clients, the sellers (restaurants, retailers, etc.) and the PDW who receives the task of having the delivery made. In very simple terms, they are the pizza delivery boy/girl of the smartphone era.

The present research is done through the combination of a digital ethnographical approach to a WhatsApp group of PDW of Barcelona and the conduction of semi-structured interviews to some members of the group.

Informed by the existing literature on precarity, my qualitative approach aims at unveiling to what extent the workers of this very specific activity are faced with a series of dilemmas from which there is no good way out -essentially they are trapped by their extremely precarious position. The interactions described in the following sections of this paper intentionally overlook any aspects that do not fall under the scope of the purpose of this research: grasping the relation between precarity and exposure to hazardous situations by PDW.

Other results notwithstanding, my expectation by doing this paper is that those whose voice is so scarcely heard, might have attention drawn to their situation, for unlike the gurus of one of most prominent companies of the platform economy (Airbnb), I don't see "people as businesses" (Kaplan & Nadler, 2016) , but rather people as people.

The deep changes brought about by the hegemony of neoliberalism as a global mantra (Harvey, 2005) produced effects on the life of all human beings on the planet, rendering the organization of societies a process in which stability and predictability were gradually, more and more, nothing but fading memories of a not-so-distant, yet apparently very far away time.

While addressing this matter, Bourdieu (Bourdieu, 1998 ) played a decisive role unveiling the "insecurity-inducing strategies" promoted by employers to render the working class more dependent and less demanding -which is synthetized by the word precarit e (precarity 1 ).

Contributions following Bourdieu's notwithstanding, Standing's Precariat (Standing, 2011) stood out as a particularly important expansion for the research on the topic of precarity. The author developed a seven-axis framework of analysis of labor security, which absence (individually considered or as a whole) led to precarity. His findings seconded those of Bourdieu, as he observed that in the post-Fordist era workers are increasingly exposed to circumstances making them precarious.

Since then, namely throughout the last decades, the fact is that this phenomenon has been greatly expanding. Not only is it evolving into an increasing share of the labor force, precarity is impacting the lives of workers in increasingly diversified ways (Armano et al., 2017) . Bauman's (Bauman, 2007) "liquid modernity" is rapidly turning into a very "age of precarity" (Lazar & Sanchez, 2019) .

In fact, the changes occurring in the labor market are deeply intertwined with further changes taking place in societal discourses and perceptions (Gilbert, 2013; Valas & Prener, 2012) . This "structural precarity" (Antunes, 2018) stems from the multidimensional nature of insecurity and vulnerability (Herrmann & Kalaycioglu, 2011) , both cause and consequence of increasing levels of inequality (Piketty, 2014) .

Unsurprisingly, the substantial technological developments of the last decade -mainly owned by the biggest capitalist corporations -have only further enhanced the critical precarization of labor relations.

"Labour power comes to be transformed into a commodity in a context where the encounter between supply and demand of work is mediated by a digital platform, and where feedback, ranking and rating systems serve purposes of managerialization and monitoring of workers." (Gandini, 2019) .

Insofar as this topic seems to be currently very trendy, which has led to rapid expansion of the existing scholarship, the in-depth descriptions provided by Srnicek (Srnicek, 2016) and Scholz (Scholz, 2017) on how these new labor forms are shaped still stand as seminal works. They shed light on how "the emergence of labor flexible forms has produced a new type of self-employed worker, one that even if autonomous in many aspects as his predecessor (schedule, low social protection) is, indeed, much closer to the temporary worker: precarious, involuntary, dependent." (Jansen, 2019) .

Building upon this legacy, De Groen et al. (De Groen et al., 2019) provide an insightful -to the best of my knowledge, the most developed to this day -account on how "platform workers" are exposed to intense and highly sophisticated forms of precarity. By scrutinizing the work and employment conditions through the lens of a four dimensional scale -the "adjusted WES" -which, in turn, unfolds into 28 different indicators, this paper stands as a landmark for a comprehensive understanding of the challenges faced by the workforce forming part of this emerging, apparently unstoppable new shape of labor relations.

All the angles from which labor relations of platform workers and their employer counterparts are looked at by the existing literature notwithstanding, the ways in which precarity penetrates platform work is far from exhausted. Among eventually others, one issue that stands outmost notably amidst the PDW -is that of platform workers renting their accounts, thereby emerging rather as subcontractors than as wageearners.

The ways in which this unfolds -to be briefly explained in following sections of this paper -has been vastly documented by the media (e.g. (BBC, 2019; Crispino, 2019; Ponte & Moya, 2019) ,), however, surprisingly, it has been overlooked by the scholarship produced by academic researchers.

As hereafter observed, the workers hired by this sort of platform's "subcontractors" are served the worst of two worlds: on the one-hand, they operate as illegal, undeclared workers, who not only are deprived any form of social protection, as also, if caught working illegally, are likely to face legal consequences; on the other hand, they are forced to pay all the taxes their sub-contractor has to pay to public institutions, and still need to comply with the control mechanisms entailed in the so-called "application-based management" (Ivanova et al., 2018) .

It stands out that this specific issue can hardly be disentangled from the vast precarity web where workers of platforms often find themselves trapped. Indeed, it should be seen more as an additional thread of the web than a stand-alone matter -more often than not, although not exclusively, closely related with situations of undocumented immigration.

This intertwined thread stands as the central axis of this paper, in which the exposure of both "regular" and "sub-hired" platform workers to the COVID-19 pandemic and subsequent societal context of unexpected and unprecedented measures of confinement is analyzed.

The constraints provided by the lockdown made it particularly challenging for any research to be conducted beyond the limits of my desk. While attempting to successfully attain the goals of myself set endeavor, I played to my advantage the increasing "ubiquitous computing" (Elwell, 2014) of human existence as an undisputed advantage.

In fact, acknowledging that the "materiality of the digital (…) should be at the forefront of theories attempting to understand contemporary practices" (Duggan, 2017) can only play to this research agenda's advantage, more so if taking for granted the quasi teleological proneness of the PDW for digital engagements (Leonardi et al., 2019) . All this rendered digital ethnography an indisputably valuable methodological approach.

Provided the term can lead to different interpretations (Dicks et al., 2005) , this is precisely what was done: while analyzing the content of the interactions of a WhatsApp group of PDW, I stood as non-participant observer of the life dynamics of this group, grasping as much as possible given the existing limitations their perceptions, concerns, attitudes and praxis.

This process was implemented from March 16 to April 12, 2020 (the first four weeks of an unprecedented lockdown for most the people 1 To be fair, the actual translation would be "precariousness". For how, to describe precarit e at labor level, one has evolved from precariousness to precarity see the Cambridge Encyclopedia of Anthropology (Cambridge Encyclopedia of, 2018).

Social Sciences & Humanities Open 2 (2020) 100059

walking the Earth), in a group to which access is publicly advertised on one of the several existing Facebook pages of PDW, and to which no prior conditions of admission are established, so there was no need for a formal introduction to the group or its moderator from my side to ensure access. However, I ended up doing it somehow, given that a big percentage of my interviewees were assembled by individually approaching each one of the group members -to whom I introduced myself as a researcher, explaining I had joined the group looking for interviewees for a research project related to the life of the PDW. This strategy also allowed me to ensure that all group members knew that there was one among them that was not a PDW, but rather a social researcher, thereby ensuring there were no ethical obstacles to my research. This was perceived as unproblematic by group members, since most of them simply ignored me, and those who actually answered (even if not made available to be interviewed) shared words of encouragement.

As for the actual analysis, as an initial step, I coded each of the 2726 entries according to the topic they referred to -a proof of how active the group, was with a daily average of 105 messages exchanged. Each category would receive one point for each entry that fell under its topic. In the not so common case that one entry would cover two different topics (e.g., requests for information on labor rhythm and police controls), both categories received a point (results are presented in the next section). The code construction was inductive, i.e., I developed it after a global content analysis of the interactions, as this allowed to a have a better look at what was going at the level of interactions -the results of this process are presented as part of the findings' section.

Although the literature review had drawn my attention to the main trademarks of precarity, such as matters of income and managerial control techniques implemented in the framework of "platform work", the general overview of interactions made it clear that more dimensions of analysis had to be explored. In line with that, it was thus necessary to address how labor precarity was closely related to other forms of broader insecurity, namely the exposure to dangers arising from either their legal status or the impossibility to control the content of the packages delivered.

Sequentially, valuable as the interactions on the WhatsApp group were, making sense of them and having more in-depth information on daily routines was only truly possible when the interviews were accomplished and successful analyzed. In other words, my rapport towards the group interactions was only built a posteriori, for it was in the interviews that the most pressing issues of the PDW lives became clear to me. This was particularly important for me as a researcher, given that I have never had the professional experience of working as a PDW, on the one hand, and that all I knew about these workers until now was essentially based on media reports describing their often precarious work and employment conditions.

In total, 8 interviews were done, all under anonymity. They were all done by videocall and, in average, lasted for 34 min. They were all conducted in Spanish, which I speak fluently as second language, and which allowed all interviewees to express themselves in their mother tongue. All interviews were recorded having the informed consent of the interviewees that the disclosure of the information retrieved would be anonymized.

The flexible nature of the interviews made them the most relevant part of this research. As hereafter observed, the information collected confirmed the validity of this technique as a means to have a wider perspective of the existing constellation of visions of the same phenomenon (Mason, 2002) .

Following Kvale's (Kvale, 1997) approach, the interviews were a mix of an unpreoccupied conversation among two acquaintances and an investigation. The script underlying these semi-structured interviews was thus quite loose, aiming mostly at having the interviewees relating their broader experience as PDW with the one of the present moment -even for cases in which they have decided to stop working, in order to protect their health. I dedicated a substantial amount of the interviews' time to understand how the work and employment conditions unfolded for the PDW, and how -in these particular circumstances -this was lived by each one of them. I focused on elements that could present themselves as hazardous, such as excessive work rhythm, overwork, absence of or difficult access to Individual Protection Equipment, lack of alternatives in face of infection, and access to social welfare mechanisms.

In both cases -those who continued to work and those who decided to stop -the interviewees' discourses shed light on the adaptations and justifications adopted by the subjects to navigate the current broader cultural framework -that of precarity.

All this said, one should not understimate the limits posed by a context in which this research process unfolded. The very strict lockdown imposed by the Spanish authorities made it completely impossible to grasp important information and its inclusion in the findings sequentially presented. This fact surely doesn't undermine the results altogether, but it has certainly prevented me from having access to broader dimensions of the precarious lives of the PDW. Those which I consider future research would benefit from grasping are: physical living conditions; social capital; mental and physical strategies to cope with the workers' immensly precarious situation. Undoubtedly, the conduction of presential (rather than merely digital) ethnography and of focus group interviews, would have made the results of the present research even deeper.

The current section presents the main findings retrieved from the combined information of the interactions in the WhatsApp group and the interviews. An insightful presentation deems useful the splitting of the information by topic, disregard their intertwined nature. A general overview of the trends of the interactions of the WhatsApp group is followed by a specific approach of the topics.

The WhatsApp group hereafter analyzed was created on January 15, 2020. It was composed of 161 members when I joined it (March 15) and kept on growing steadily until the last day of my research (April 12), when it counted with 176 participants -however not all participants were active, many of them stood as mere "observers". The group founder and moderator, himself also a PDW (assuming his profile photo is actually of himself), was one of such silent members -not even once has he interacted publicly within the group. The participants were mainly PDW who worked in Barcelona, although there were a very small number of exceptions -people from other Spanish cities looking for info or technical information.

Although most of the interactions referred to Glovo, there were PDW who worked for other companies such as Deliveroo, Uber Eats and Stuart. Some PDW had accounts open in more than one platform.

Elements like age and country of origin are impossible to account for. However, admitting that the photo provide by the profile picture is, most of the times, from the owner of the WhatsApp identity, it was from the beginning clear that this group where young adults are the overwhelming majority (something that was later confirmed by the pool of interviewees, whose age ranged from 21 to 33).

The same rationale applies for gender. Based on the WhatsApp profile photos and on the usernames -admitting they were reliable -I could only count seven women. Notably, one was a fellow social researcher performing digital ethnography as I was doing, another was the wife of a PDW (it was unclear whether she, herself, was also a PDW) -the remaining others were actual PDW. This does not mean that all other members of the group were men: in the case of at least 32 members, neither their usernames, nor their photos allowed to make a solid prediction on their gender. Be as it may, during the two female interviewees confirmed that women were extremely rare in the PDW world.

As for the content of the interactions, after a global reading, I deemed to code the information retrieved from the group in the following manner:

COMPANY -dealing with aspects of the relation with the companies (mainly doubts and complaints shared among peers, before/instead of addressing the company itself); LABOR RHYTHM -related to the ongoing movement of demands made by clients (usually to assess if low levels of requests were an individual situation or a broader issue); POLICE/SECURITY -shared information about (past or present) police controls and other aspects that could jeopardize the security of PDW, such as requests for delivery of illicit substances; RENTED ACCOUNTS -information on accounts to be rented (essentially from people procuring them, but also some advertising and some doubts' clarification); STATE -discussion all aspects related to the relation of PDW with the state, mainly: the decision of the state of emergency, its renewal and its implications, and clarification of doubts on issues related with taxes and social security; COVID -information and debate of news and concerns directly related to COVID-19 (the evolution of numbers, who is to blame, what measures to take to reduce of the possibility of being infected); SERVICES -advertisement, sharing or procurement of services, disregard how work-related they are: second-hand backpacks, bikes' workshops, flats for rent, cigarettes, fuel, videogames or even direct requests to workers seeking to avoid the fees charged by the apps; OTHER -jokes, memes, motivational messages of various sorts.

As shown by Fig. 1 , throughout this period, the main concern of the DPW was related with the action of the police and other aspects related to security, taking up almost one third of all the interactions. Most notably, concerns with COVID-19 itself were extremely rare (2%).

In fact, direct work-related worries were by far the most pressing, adding up to an aggregated result of 89% of the interactions. Even if the Services category is not considered as necessarily work-related -although a large number of the interactions that form part of it indeed are -, one would still observe that 74% of interactions refer to the need to deal with aspects that are connected to work.

The "Others'" category -which altogether added up to 9% of the interactions -was the locus of empathy among the PDW par excellence. Inasmuch as these were not directly work-related contents, it was clear that group members found in this type of message a way to mutually motivate one another, expressing individual and collective resilience before such complex context as that of the COVID-19. Undoubtedly, most messages were or religious (or related to them), even if, seldom, some jokes, memes and gifs would be posted, particularly as reaction to some previous message, as a way to provide feedback. 2 The rest of this section is highly informative to why this is and how each of the different dimensions is, one way or another, related to the specific circumstances brought up by the COVID-19 pandemic.

If there is one thing that seems consensual among the interviewees is that working as PDW provides a low monthly income. Most have noted that work is exhausting, allowing to obtain an amount with which one can "pay your food, your rent, and not more" (Juan). Of those interviewed the declared net income ranged from 300 to 500 euro per fortnight, all highlighting that reaching these amounts deemed necessary a very intense work rhythm, with no more than one day per week to rest.

An important element of instability stems also from the inexistence of a minimum guaranteed wage, rendering it pure luck that there will be enough movement to allow one to earn enough to pay their bills. This is arguably why almost 10% of the interactions group are dedicated to understanding how rhythm of demands is at each day. The randomness of this possibility to make money is well expressed by the following statement of a member of the WhatsApp group:

April 5: WG#1 "One has to be patient, that i learnt from [name of company] … One just has to get out soon and have faith in God that a request drops."

Furthermore, the situation of the PDW is of people who live on the edge, renting rooms by the day, with no security at all, exposed to the arbitrariness of those with more power -something that became even more critical during the pandemic, leading one to believe that the COVID-19 crisis may just be the tip of the iceberg of much deeper crisis, where job, food, shelter and medicines' scarcity may lead up to a not so distant perfect storm. Take as examples: 2 Although the jokes and gags could, altogether, be a strand of this research on its own -following the steps of, among others, 't Hart (t Hart, 2007) -their scarcity and discontinuity rendered this possibility unattainable.

March 22: WG#2 "I could not go out to work because the person who is renting me [a room], told me to look for another place to live, that they did not want to expose to the possibility of me taking the virus to the flat …"

March 29: WG#3 "If you know of any bicycle account recommend it friends, maybe having some money it is easier to find a place to sleep, this night one good person let spend the night so I don't get frozen with the cold of the last days. I have a bike to work. If you know of any account please I'm very unsettled"

Beyond income, a decisive element to understand the whole context lived by most PDW is the way the apps are run, commonly known as application-based management.

Although each platform (which tangible expression is the application) has its specific features, the outline is similar for all. In a nutshell, the odds of a PDW having possibility to work (and, necessarily, obtaining income) is directly related to a mathematical combination of factors, such as: the availability to work in the hours deemed necessary by the platform (most notably in the evenings and weekends), the speed of delivery, the review the given by the customer who gets the order (and, in case it is food, also from the restaurant producing the food).

When accomplished within the goals set by the platform, the PDW obtain points. When that is not the case -no matter whose fault it is -the PDW loses points. This extremely important since it is the amount of points one has that determine one's ability to "catch" hours of work. The more points one has, the more hours one will be able to reserve -hence, the more money one can make.

Diego pointed out how the lockdown was a special period, since "In this period they let you choose the hours you want, something that never happened", which stems from the huge demands received in these days, particularly for food orders. Juan reported that the company he delivers for was now increasing the income rate to attract workers to work for more hours, something that had been done in the past only in rainy days or special holidays.

Be as it may, there still are the "hours of high demand" (defined by the platform and the restaurants), in which "if you don't work, your score will be lowered" (Juan).

In a context like that emerging from the COVID-19 pandemic, it should come as no surprise that many people have tried to desperately find solutions to economically survive. In the month of March alone, more than 900 thousands of workers were affected either by unemployment or by their companies making use of lay-off mechanisms (Gomez, 2020) . One way or another, the financial stability of hundreds of thousands of families was directly affected, rendering them more prone to searching solutions where there might be potential for fresh and fast income.

If one goes back to Fig. 1 , the fact that more than 1 in every 10 interactions are related to this topic provides evidence of its relevance. However, to fully grasp the meaning of this number it is important to keep in mind that these are interactions about performing an illegal action held in an open forum, in which everyone can be easily traced by their mobile number. This renders highly likely that many more interactions take place in non-public spaces.

From the reports in the media and the interviews had, it is relatively easy to understand how this unfolds. In a nutshell: Individual A signs-up for working in one platform, providing the required paperwork to be granted a license; once that is obtained, Individual A procures someone who is willing to work but cannot/is not willing to sign-up for working in the platform; once that person is found (hereafter Individual B), an oral agreement is celebrated: Individual A grants access to his/her account to Individual B, provided that the latter concedes the former a percentage of the income earned by working to the platform, plus all taxes. In all cases, Individual A is in control at all times: the bank account where the money is deposited by the platform is his/hers; the mechanisms for password change are in his/her hand.

The situation reported by Jos e speaks for itself in terms of the vulnerability to which people renting accounts are exposed:

"In of those payments he decided not to give me a thing, not to pay me … making excuses, that he wasn't making any money with that account and I don't know what else … and he stole that fortnight from me. That was a great problem, I had to pay the rent, and to eat also and, well, that time we even had to put up with hunger, because we lived very poorly …"

A side, but important note to take here is: Traditionally, this is a situation in which undocumented migrants are more likely to be found 3both the existing literature and the interviews made confirm it. However, in the pandemic context, it is likely that Spanish nationals and documented migrants are drawn by the perks of renting an account: skip all the (although light, time consuming) bureaucracies to become a legal PDW and start making money "right away". 4

While some companies only deal with distribution of food, there are others who provide delivery of packages of any sort (as long as they fit in the backpacks of the workers), including items between two addresses of two private individuals.

The second half of this paper's research period was marked by the emergence of cases related to drug traffickers making use of these platforms' services to have their products distributed around the city -thus avoiding having themselves or their customers exposed by walking around the highly patrolled streets of Barcelona.

An example:

April 6: WG #8 "Well not long ago, a super beautiful blondie asked me to take a box of chocolates … and when she passed it onto my hands … and weighted nothing … that made me curious … And when was coming in the lift I found it suspicious and I felt like smelling it, and it was pure marijuana in the box of chocolates …"

Unsurprisingly, the lockdown was enforced by the action of police. Practically empty, the streets and avenues of Barcelona became constantly patrolled by police officers enquiring the purposes of people going around, no matter if they were on foot, riding their bikes or cars. 5 Allowed by the governmental decision on "essential activities" to keep on working, the PDW were not unexposed to trouble with the authorities -reason for which 30% of the interactions on the WhatsApp group was about the police controls. As expected, the topic also came up 3 In which case the account owner has an additional power over the individual renting it: that of, at any moment, denouncing him/her to the authorities. 4 Those renting an account for this purpose will not be caring for the last, but not least problem stemming from running a rented account: the absence of any form of social protection or benefit. This became outstandingly important when the Spanish government announced a series of measures to support the selfemployed workers (which, in theory, is the legal category of the PDW) and those running rented accounts could not enroll for having access to such mechanisms. This was, by far, what occupied the 12% of the interactions above categorized as "State". 5 In the first two weeks alone, there were more than 1.500 detentions and 180.000 fines for unjustified violation of the lockdown in the whole of Spain. Notably, from that the day on it is not possible to find overall data.

in most of the interviews, although not with such relevance. Indeed, the three reasons behind the concerns with the police controls are deeply intertwined with the previous four dimensions of precarity. Even if, in most cases, an overlapping of issues is observed, they are hereafter split for a more comprehensive description.

The systematic disrespect for traffic rules has very few (if anything) to do with the COVID-19 context, however the huge reduction in traffic and having "police coming out of every corner" (March 27, WG #11) made the situation more hazardous.

As noted by some, even when not fined, a PDW will necessarily be doubly harmed by a police control: the time spent while having his backpack and documentation checked will necessarily lose him precious hours of work and, simultaneously, impact the evaluation given by the client who will receive his package later than expected -something particularly critical in the case of food coming from restaurants.

The words of Jos e provide a clear picture of why PDW are so eager to avoid the police:

"I cross a lot of red lights. And why do I do it? One is working with the hours given by [name of company], so one has to do what is possible to answer more requests in that hour, and so one goes as fast as possible to work more. Sometimes I am afraid, because as much as one is careful, there is always risk" B) Rented accounts Although they are paying taxes to the account owner (rendering this slightly different from other forms of undeclared work), renting an account is an illegal action and, as noted above, it is often performed by undocumented immigrants. This deems police controls particularly unwelcome for many PDW, with consequences ranging from fines to severe legal complications.

This widespread shared concern informs much of the interactions in the WhatsApp group, as well as it emerges in several interviews. It is an undisputed source of anxiety:

Juan: "Now they are also asking for documentation, and I will present the documents of my country and they will see that the account is someone else's … and things get complicated …"

March 16: WG #12 "If one has a rented account, and besides it one doesn't have a passport, then you will have to take a lot of precaution, take latex gloves if possible, mask, lights on the bike, that way we avoid drawing attention." C) Transport of illicit substances. It goes without saying that no one will want to be found by the police carrying illicit substances. Scrutinizing the information retrieved from the WhatsApp group (by the times the interviews took place this wasn't yet an issue), it stands clear that the PDW faced an unprecedented situation, which rendered many of them anxious.

The discussion floated around how to avoid being used by drug dealers as carriers, and simultaneously, how to act in order to ensure the police would believe that -when they did carry drugs -they belonged to the client, and the PDW was unaware of what was being carried:

April 11: WG#16 "Guys, the best is not to take packages, there are already many cases of these, I know we are here to earn money but we can't risk, so even if we take a thousand photos, if the police will stop us, it is enough that it is in our package to have us being the harmed ones"

Against all odds, the issue of the COVID-19 deserved very scarce attention from the PDW, particularly in the WhatsApp group (only 2% of the interactions were directly related with it!). In fact, most of the references to it were purely instrumental and referred to how it connected with other dimensions of the PDW life, namely whether the state of emergency would interfere with their possibility to keep working.

This statement should not, however, overlook the direct impact that COVID-19 had on the life of several PDW. For instance, afraid of becoming sick Jos e and Maria reported having stopped working, while Juan spoke of being extra careful -not afraid of the disease itself, but rather of, by becoming sick, not being able to work and, therefore, having no means of survival.

Despite different approaches to the topic, one aspect seemed to be consensual: the answer of the corporations that own the platforms was seem as scarce and perceived with disbelief by the PDW:

March 29: WG #17 "They don't give you even a glass of water, let alone a mask. I came from there. And they don't even thank you" WG #18 "I saw some guy asking for it and they give them [masks], as well as gloves. Obviously, when they run out, they won't give you … And you have to ask for it, it's not like they will be announcing it … Their obligation would be to even notify us through the app, but we already know how it is …"

To a large extent, the pandemic brought up the already existing precarious situation of the PDW. Although the pool of interviewees does not stand as a representative sample -nor was that ever the intention -, it is clear that the PDW are mass of people who could not make their way into the world of standardized wage-earning decent jobs. Either due to their situation of undocumented migrant, or simply because youth unemployment is massive in Spain, working as a PDW forms part of the "precarity trap" in which millions of workers increasingly find themselves (Fumagalli, 2013) .

Having this in mind, it is of the essence to the latter understanding of the emergence of a set of dilemas that, in a different context, could have assumed substantial different shapes, or not have surfaced at all.

As aforementioned, the interviewees and the interactions in the WhatsApp group made clear that there were different levels of concern regarding the possibility of being infected (and possibly killed) by the COVID-19. Nonetheless, it stood crystal clear that, even for those exhibiting less fear, the decisive driver of their decision to work (in the case of those who did) was their impossibility to have reach out to any sort of alternative whatsoever.

As per the definition of the Merriam-Webster Dictionary ("Dilemma." Merriam-Webste, 2020) a dilemma is "a usually undesirable or unpleasant choice". Provided the above-mentioned circumstances in which PDW find themselves, I consider that, in the present context, they face 4 critical dilemmas -which are hereafter systematically exposed:

By far the most critical of all, the survival dilemma stands as a trademark of this particular time. On the one hand, by working and contacting dozens of other people (restaurant workers, clients, police officers, random people on the street), PDW are highly exposed to the highly contagious COVID-19. On the other hand, the absence of mechanisms of social support (such as unemployment or health benefits) and the possibility of having their account renting agreement unilaterally cancelled (for not being as productive as the original owners of the accounts expect/demand), impels them to continue work.

Over the period of research, I found two highly informative accounts, which I believe to efficiently summarize this dilemma:

Pablo: "Of course there is a widespread fear … but even more fearsome is not to eat or having to stay on the street. You may get the virus and heal when it's over, but if you don't have anything to eat, you or your family, the death is of starvation. So, the fear is from hunger and that's it. I mean, the delivery workers, we are not fighting against one virus, we are fighting against several."

March 18: WG #19 "I understand all that, but i have my daughter and the rent won't be forgiven. And my fear is to bring the virus to my house, to my wife and daughter. It is being between the sword and the wall. I'm desperate." 4.11. The cost minimization dilemma: fines versus absence of income Particularly for those not having an account under their name, the existence of a higher level of police control in the streets raises yet another dilemma. While going out is decisive to earn income to pay for their basic expenses, PDW caught using other people's accounts without documentation risks being severely fined -which would, of course, render completely useless all the efforts for gathering money.

Jos e: "I decided not to go out because this is a serious problem. What's more, I heard that the police was in the streets asking the delivery workers to show their working license and the social security documentation. From my side, I have no way to deliver that because my account is rented, so there is a risk I'm fined, and I've been told it can go up to 600 euro. This fortnight I haven't earned a single euro."

April 5: WG #20 "Partners i have a rented account and I am about to run out of euros, what do you recommend, get out or can I get even more screwed with a fine?" WG #21 "With the fine or with virus, any of the two." WG #22 "Of course you can get more screwed but you have to work. Get out from 12 to 5. Whatever you do then is good. I go out everyday and I'm like that" 4.12. The algorithmic dilemma: red signs versus police attention

The third dilemma is not specific to this period, however the ubiquitous presence of the police in the streets makes it rather more pressing than in the recent past.

The quotes provided in previous sections stand as the most significant to shed light on this issue, reason for which no new quotes are provided here. Nonetheless, this dilemma's underlying rationale is the following:

On the one hand, the only way to "please the algorithm" which lies at the foundation of the app is to work with high pace, since that will improve the odds of having good reviews, which, in turn, will allow the PDW to have more hours available to work and, thus, earn more income. Even if the results of a high rhythm of deliveries doesn't have immediate algorithmic impact -although far from only long term, since the score of each PDW is updated on a daily basis -, as noted by some of the interviewees, the fact that the PDW get one delivery done, allow them to move on to the next one within the available hours. This will be particularly critical to those start to work as PDW, since they will have both fewer hours available and, within those, fewer odds of having a delivery requested to them. Notably, this seems to confirm that technology is not, on its own, a form of control or exploitation: the ways in which it is designed and implemented rather mirrors the social relations it embodies (Moore et al., 2018) In the case of the PDW, the absence of human (visible) interference in the decision-making process, allows managerial control to be perceived as computational and, thus, non-biased towards the boss or the employee (Gillespie, 2013) , thus impelling workers to incur in hazardous behaviors, although keeping the appearance (in some cases some sort of fetishicized illusion) that it was their choice all alongand not the companies' perverse algorithmic design inducing them to adopt such praxis (Veen et al., 2019) .

On the other hand, to ensure that this very high rhythm is accomplished demands to not always act according to the laws, particularly traffic rules. Such an attitude may, however, entail a series of negative consequences, all of them steaming from having the police attention drawn to oneself. The first, and most obvious, is that being stopped by the police will mean losing time, thus rendering the risks and the efforts to make a fast delivery useless. The second, almost as likely in case is not respecting traffic rules, is that one may be fined, which will obviously impact on the available income as soon as the fine is paid. Third and most seriously, particularly for those using rented accounts, having the police checking their situation will probably lead to even heavier sanctions, ranging from bigger fines to serious legal complications (particularly for undocumented immigrants, which, as above mentioned, is many times the case).

Ironically enough, given the more precarious one's situation isspecifically in the case of account renters -, the more one will feel compelled to violate the laws, given that the net income retrieved from working as PDW will be even lower.

As noted in the previous section, a problem that has been emerging for the PDW is that there seems to be a sudden increase in cases of drug traffickers making use of them to ensure the distribution of their product. Such fact poses a series of problems that dialogue directly with several critical features of the PDW activity and the way it is designed.

On the one hand, PDW afraid and/or suspicious of being used as involuntary carriers of drugs may choose to take one of three different actions, all of which may have important consequences:

(i) Refusing the request for deliver packages from private senders: the immediate consequence is that the earnings that would be obtained from that service will be missed, without the PDW knowing when will he/she be able to have another service; beside that, rejecting services is very likely to have negative impact on one's score, thus affecting his/hers possibility to earn money in the future. (ii) Requesting the sender to open the package to ensure it's drugsfree: by asking something that does not comply with any recommendation of the platforms and is, most likely, illegal, PDW risk being sanctioned, either by having a bad review and/or if denounced by the sender to the main company; furthermore, even if this is not at all negatively interpreted by the sender, the fact that the PDW is losing time will already stand as an indirect sanction, for as noted in the previous dilemma it will likely harm the total income earned. (iii) Voluntarily asking the police to check a package: particularly under the ongoing circumstances, the police has the authority to open a package and provide a legal document that the PDW may attach to it, thus defending him/herself from any accusation of violation of privacy, thus rendering this the safest option; however, this option will make the PDW lose time of eventually scarce available hours to work, which will have the impacts described above; furthermore, for an account renter (let alone an undocumented migrant!), voluntarily stop at a police control will most probably mean having documentation checked, which, for what has been exposed previously, will hardly stand as a true option

On the other hand, not taking any of these three actions poses a serious risk of being stopped by the police anyway, except in this scenario, beyond any other eventual illegalities performed, carrying drugs (or even throwing it away, as some proposed in the WhatsApp group). While the PDW may argue that this was being done as part of the job, proving it may be far more complicated -notably, this will be particularly more complex for undocumented immigrants.

A relevant excerpt would be:

April 11: WG #23 "How heavy it is with this thing of the deliveries. Even if you reject it because you know its drugs and report it, you lose points …"

WG #24 "My husband asked the client to open the boxes in front of him and the man didn't want to. He obviously didn't take it. And told him all sorts of things but his security is first, since he is already risking by going out because he rents an account. Of course that for the police you are an accomplice and that's it."

All things considered, exceptions notwithstanding, the PDW options are everything but true choices. The insecurity to which they are exposed deems necessary to incur in ever-growing risks (inter alia, exposure to disease and security risks, high degree of probability of being fined by of the police, or even worse) -which renders these as practically false dilemmas: at the end of the day, there is no choice but to comply with the companies' rules and go to work.

As I write the last lines of this paper the developments of the COVID-19 are still very much unpredictable. Be that as it may, out of the many conclusions that can already be extracted from this health crisis -and which may prove of the essence for the more than likely upcoming economic crisis -, one is surely that there is no such thing as everyone being on the same boat.

If anything, the COVID-19 crisis has exposed the multiple-fold consequences of precarity. Not that the immunity of some human beings is stronger than others just because of their job and/or economic security, however, the exposure to health and other risks, the way one cope's with temporary economic scarcity, the very ways in which one goes by a period such as this (for now) one month of quarantine, are not the same for everybody.

As part of his Precariat, Guy Standing (Standing, 2011) drew attention to how the expansion of precarity has meant the reemergence of the ancient denizenry, i.e. a mass of people whose absence of many rights denied them the entitlement to a full citizenship. While this has always been the case of vulnerable layers of society, such as the undocumented migrants, as this paper has confirmed, these are far from the only ones falling onto this dark category.

Inasmuch as the existing constraints to circulation and the rapid unfold of events posed some limitations to the elaboration of this papernamely the possibility to have a larger pool of interviewees -, it still stands as a contribution for a broader understanding that the modernization of the labor process does not necessarily entail progress or social justice.

Nonetheless, still further research on how the COVID-19 crisis impacted the life of precarious layers of societies is needed. Among others, two critical aspects stand out as not possible to have been fully grasped by the research hereby entailed: i) To what extent the PDW were exposed to health risks to ensure they would fulfill their mission, and what did the corporations they work for do to actually tackle those risks. Ii) Disregard the public discourses surrounding the PDW's importance, in a moment when many people are forced (by law and/or by risk) to not to leave their homes, at least one of the companies -most notably the biggest in Spain -has decided to lower the retribution per delivery ("Dilemma." Merriam-Webste, 2020); beyond the outstanding contradiction, both the justifications and consequences of this decision remain unpredictable, but of the utmost interest from a sociological standpoint.

As a final word, the present paper seems to underline that unless policies are designed to reconcile the development of technology, justice and human development altogether, the odds are that groundbreaking innovations are appropriated by those already concentrating wealthusing them to their advantage to further improve their already privileged situation -even if that means voting those around them to indecent living conditions.

If ever there were any doubts this is the way unregulated capitalism works, the situation of the PDW in the age of COVID-19 leaves very few room for doubts any sort.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement Tiago Vieira: Conceptualization, Methodology, Software, Data curation, Writing -original draft, Visualization, Investigation, Validation, Supervision, Writing -review & editing.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Humour and social protest: An introduction

O Privil egio da Servidão

Introduction

Vida líquida

O mercado ilegal das contas de app de entrega alugadas por brasileiros em Londres

Acts of resistance: Against the new myths of our time

Caporalato digitale tra rider, account italiani venduti a migranti irregolari: «Dammi il 20% e ti cedo l'account

Study to gather evidence on the working conditions of platform workers

Qualitative research and hypermedia: Ethnography for the digital age

Questioning "digital ethnography

The transmediated self: Life between the digital and the analog

Cognitive biocapitalism, the precarity trap, and basic income: Postcrisis perspectives

Labour process theory and the gig economy

Common ground

The relevance of algorithms

La crisis del coronavirus provoca la p erdida de 900.000 empleos desde el inicio del estado de alarma. El País

A brief history of neoliberalism

Precarity -more than a challenge of social security or: Cynicism of EU's concept of economic freedom

The app as a boss? Control and autonomy in application-based management. Arbeit | grenze | fluss -work in progress

Self-employment as atypical or autonomous work: Diverging effects on political orientations

Airbnb: Case study in occupancy regulation and taxation

Det kvalitative forskningsintervju

Understanding labour politics in an age of precarity

The ambivalence of logistical connectivity: A co-research with foodora riders. Work organisation

Qualitative researching

Humans and machines at work. Monitoring, surveillance and automation in contemporary capitalism

Capital in the twenty-first century

El mercado negro de los «riders» de Glovo en madrid: Alquilan sus licencias a «sin papeles

Uberworked and underpaid: How workers are disrupting the digital economy

Platform capitalism

Dualism, job polarization, and the social construction of precarious work

Platform capital's 'app-etite' for control: A labour process analysis of food-delivery work in Australia

This paper would not have been possible with the insights and incentive of Mireia Bolíbar, without the comments of Daniel Romero-Portillo and Adam Tash, and without the contributions and companionship of my quarantine and life partner, Ana Oliveira. All of them deserve may utmost gratitude.