From left to right: (1) a screenshot of Uber Driver app containing the representation of surge pricing in São Paulo, the Brazilian city with the highest occurrence Covid-19. posted by a driver on a Facebook grupo on May 11th; (2) A Screenshot of the Covid-19 confirmed cases map produced by

[BigDataSur-COVID] Managing uncertainty at your own expense: Brazilian Uber drivers during the Covid-19 pandemics

With Brazil becoming the hotspot for the COVID-19 pandemic, Uber drivers in the country have to choose between putting their wellbeing at risk and giving up their main source of income. But is there anything new about this? We spoke to three Uber drivers in Belo Horizonte to find out.


By Ana Guerra

Leia en Português

Brazil is Uber’s largest market outside the United State, counting over one million “partners”. Recently the country has also marked another record as the world’s worst COVID-19 outbreak after the United States, with 772,416 confirmed cases and over 39,000 deaths as of June 11. This article explores the impact of the pandemic on a population at the (socio-economic) margins, namely Uber drivers. It gives voice to their concerns as they roam the streets of Belo Horizonte, the 6th most populous city in Brazil.

Inside Uber’s predictive modeling 

The transnational ride-hailing platform operating in 63 national markets around the world is a known example of on-demand economy, or “gig economy”. While affording a relative amount of independence to workers, it provides no stability or safety nets. Working under a highly datafied labor regime, drivers find themselves caught up in exhaustive routines and short of labor rights, since they are not classified as employers.

How does such a model work? Uber employs a series of techniques to algorithmically manage driver’s labor in global scale, often presented as tools to predict supply and demand, stabilize uncertainties and help drivers make more profitable decisions. This is made possible by the platform’s continuous collection and processing of data about traffic patterns and the rider location and behavior. The company’s developmental and rhetorical efforts alike make apparent the ambition to manage and mitigate uncertainties. This is illustrated, for example, by the platform’s forecasting solutions aimed at predicting and managing the spatiotemporal dynamics of what is referred to as “the real world” on Uber’s blogs, as well as by its patent applications. The desire to anticipate future dynamics through data is also made evident by Uber drivers’ “algorithmic labor”, as Rosenblat and Stark call it.

Uber is a paradigmatic case of the precarization of labor through what is commonly known as “gig economy” or “sharing economy”. In a way, it is no surprise that this process is often referred to as “uberisation of work” ( “uberização do trabalho” in Portuguese). The incorporation of predictive modelling to workers’ routine reverberates what Adrian Mackenzie has identified as the generalization of prediction into everyday life, meaning the increasing presence of processes such as the algorithmic sorting of preferences and recommendations, pattern finding, demand forecasting, and other resources aimed at stabilizing future outcomes and actions.

The dangerous cocktail of datafied performances, job insecurity and poverty is particularly toxic for drivers during a pandemic, as characterized as it is by uncertainty hitting all spheres of human life.

Managing uncertainty in a pandemic

With COVID-19 taking over our thoughts and affects, the real world seems to be getting a lot more real and our relation to prediction a little more intimate. The world as we know is rapidly transforming into something else and sometimes it is hard to keep up with the numbers that tell us what happened today and what to expect for tomorrow: how many new cases? How many deaths? How many days left until things go back to “normal”? The answers change day after day. The new coronavirus twisted our experience of time and space. An invisible agent, which we may or may not be carrying around inside our cells or on our clothes, rapidly unsettled the steady references around which we organized our routines. As a wide variety of data, predictive models, projections and visual representations try to render the situation a little more intelligible, the unknown keeps finding ways to confront us from unexpected angles. Any flash of certainty is found to be short lived and no amount of projections or animated graphics are able to minimize the problem. In a way, our “dataistic” narrative that privileges datafied and prediction-oriented ideals of truth is progressively destabilized. The notion of “dataism”, as José van Dijck  explains, speaks of the belief in the “objectivity of quantification” powered by the datafication of human behavior and sociality on digital media platforms. Yes, we keep on counting— counting patients, deaths, days — but our pace has changed.

Despite the globalized sense of uncertainty, some places feel more uncertain than others. In Brazil, the COVID-19’s next hotspot, the “real world” is met with a pint of fantastic realism. Amidst underreporting, lack of testing, endless conflicts between health authorities’ and President Jair Bolsonaro positions, news about collective graves and fake news about caskets filled with rocks, Brazilians find themselves in a rather chaotic scenario. Meanwhile death rates keep rising as the poorest layer of the population is increasingly affected by the virus and the collapsing social-economic system.

Among those caught into despair are many of Brazil’s Uber drivers. Since Uber entered Brazil in 2014, the platform quickly positioned itself as a transportation solution for lower income and underserved communities, forging a quasi-infrastructure role. It goes where many taxi drivers will not go and where public transportation coverage is rather deficient—like favelas (slums in English) and urban peripheries. Besides, an Uber ride may be just as cheap as taking the bus and is typically much faster. Uber also holds a privileged and ambiguous position as an unemployment savior, since many Brazilians rely on Uber and similar platforms as their major source of income. Right now, however, drivers are forced to decelerate. Some of them report up to a 90 percent decrease in rides.

Approaching uncertain times: Uber’s response vs. driver experience

Uber’s responses to the pandemics do little to appease uncertainty. Among a pack of “resources” directed to Brazilian drivers, the platform announced an up to 14 days “financial assistance” plan to “partners” either diagnosed with COVID-19 or classified as suspicious cases or part of a at-risk group. Their eligibility must be attested by official documentation containing detailed health information. Rather than offering a fixed common value, the assistance is calculated based on the driver’s average earnings for the previous three months, being closely tied to their individual performance. The policies that inform the assistance distribution seem to be short-run, and the platform itself indicates they may be often updated. The rules described here were updated on April 17 and are said to be valid only until May 8 but the deadline was later updated to June 8, 2020. This is exasperated by a paradoxical treatment of “risk”. Once drivers request assistance, their accounts are automatically deactivated for safety reasons. However, this does not guarantee their eligibility, putting them in a dubious position: the driver poses as a risk, and therefore is prevented from working, while simultaneously not being at risk—thus, no assistance is granted.

To learn more about how Uber drivers approach this conundrum, I interviewed Giacomo Antônio, and Verón three drivers operating in Belo Horizonte, a populous city of over 2.5 million people and the capital of the state of Minas Gerais. I also analyzed over 50 comments to a YouTube community post by Samuel, an Uber driver and YouTuber who shared my questions with his followers. At the time of writing, all three interviewees are still working. As they drive, SARS-COV-2 may be riding in their back seat, a risk worsened by careless riders. Giacomo estimates that out of ten passengers he picked up, only two wore a mask. Antônio’s wife is diabetic, a risk group for COVID-19. He however chose to continue working so they could afford her the balanced diet she needs. He purchased hand sanitizers, and even though he could request Uber’s one-time 20,00 Brazilian Real (BRL, approximately EUR 3.5) refund for hygiene items, he chose not to because he considers it a “derisory help”. “It’s offensive”, he says. The three drivers share the sensation that movement has slowed down, which might mean sitting alone in the car for a couple of hours, waiting (and thus not earning). The job became a little more solitary. After his son was murdered at 19, Verón put his life on hold for over two years, “and then I met the apps”. Since then, he sees driving for Uber as a sort of therapy and an opportunity to meet new people. He notices that since pandemic started, people became less inclined to chat during rides.

Most drivers who answered to Samuel’s post seem to have quit driving for the time being. Fear and safety appear as the main reason for that: “I already had a sense of fear before COVID-19, let alone now,” one driver wrote. Working or not, drivers readily felt the financial consequences of the pandemic. About 160,000 drivers, according to car rental companies, decided to return their rented cars. Others are unsure of how they will pay the next installment for the vehicle they bought precisely to become Uber drivers. Many resort to the emergency income provided by the government, worth BRL 600/month, approximately EUR 107). While the people I spoke to had some requests accepted and a couple declined, most drivers are met with the “under evaluation” message accompanied by an ambiguous advice to “try again tomorrow”. Some have been trying again for over a month. As for Uber’s assistance, the majority of drivers do not meet the requirements and exhibit a general feeling of distrust towards the platform. “Uber does nothing for us,” argued a driver.

But is there anything really new under the sun?

Asked to describe the current moment in one word, drivers drift from “uncertainty”, “fear” and “frustration”, to “resilience” and “perseverance”. But what does this mean? Are those new feelings? Looking at how the pandemic affected Uber drivers’ experiences may be more prolific for what it renders visible than for what it gives rise to. These circumstances invite us to defamiliarize a recently reinvented, but long naturalized, state of precarity. As philosopher Judith Butler suggests, by asking about Uber drivers conditions under the pandemic, “we are also asking about the conditions of life and death that hold for the social organization of labor”.

Uncertainty is no novelty. As much as they are used to being counted—working under a highly datafied labor regime, continuously being held up against their performance metrics—Uber drivers are also used to count. Trying to estimate their daily earnings while dealing with the variable fee charged by Uber after each ride, calculating their expenses for gas and maintenance, planning to pay for car rent or installment, are all part of drivers’ routine. Drivers also spend time and energy coming up with strategies and goals to optimize their productivity. Life as an Uber driver is marked by a precarious drive towards short-term predictability, while the future remains largely obscure.

If we bring risk back into the equation, the conditions of life and death are made more evident. Butler asks “who risks their lives as they work? Who gets worked to death?”. Getting worked to death is a common metaphor for Uber drivers. As a driver I interviewed back in 2018 put it, after a car damage prevented him from working for over 20 days, “now is the time I work to wear myself out, for as long as my body can take it”. Sadly, the reference to death goes well beyond the metaphor. Fear of violence is a strong component of the shared experience of drivers in Brazil, as they find themselves vulnerable to robberies, kidnapping and murders. It is not uncommon to read news about an Uber driver’s body being found a while after he or she was reported missing.

Thus, while Uber’s business model and technological development revolves around mitigating uncertainty through data processing and predictive modelling, it turns out drivers are all too used to managing uncertainty at their own expense. The difference lies in scale. Uber wants to stabilize the real world, predict demand, manage labor and rationalized urban traffic dynamics across the globe. For drivers, uncertainty hits much closer: it is about whether they will make it through the month, or even to the next day. It is about feelings of fear, despair, hope, resilience, pride and tiredness.

Drivers are well aware of how precarity constitutes their daily routines and fight to make a difference. Better fares and more safety have been their main demands for a while. For now, they spot no signs of better conditions sparkling when things go back to “normal”. Some argue this is actually the right time for drivers to get their voices heard—if only people would stop “risking their lives for handouts”, laments a driver under Samuel’s video. Mostly, as uncertainty keeps rising, a sense of urgency seems to prevail. I ask Giacomo about the main needs of drivers during the pandemic. He promptly replies: “what we need is rides. We need rides”.


About the author

Ana Guerra is a Masters student at Universidade Federal de Minas Gerais, Brazil. She currently researches platform and algorithmic labor in Brazil, with an emphasis on Uber and the experiences of Uber drivers. Ana can be found on Twitter @anagvguerra.