Aiming to contribute to the understanding of how the relationship between autonomy and control is materialised in ‘work by app’, this article draws on the results of research conducted in the city of São Paulo, Brazil, with Uber drivers, The research aimed to analyse the management practices in ‘app’ work. Because this survey was conducted in the global South, we also present some considerations about the Brazilian working-class conformation and the conditions under which work by app is being performed in Brazil. We argue that the algorithmic management made possible by adopting apps across labour processes is consolidating a new form of management, organisation and control of labour power, increasing workers' actual subsumption to capital and radicalising forms of work exploitation and domination.
The phrase ‘work by app’ is used in this article as a synonym for ‘platform labour’. We use this term because it is the best-known terminology among the Brazilian population. Platforms are privately owned or publicly owned data-driven software and hardware infrastructures that are automated and organised using digital algorithms (Casilli & Posada, 2019), enabling interaction between two or more people or groups (Srnicek, 2016:43).
The data used in this article are the result of 22 interviews, with semi-structured script, conducted with drivers and former drivers who provided services using the Uber application in the metropolitan region of São Paulo, between 2018 and 2019, twelve of which were conducted during the Global App Driver Strike, which took place on May 8, 2019 and brought together approximately 200 drivers in the downtown area.
Fuchs & Sandoval (2014) argue that there are 1,728 different possible forms of digital work, ranging from ore mining activities for laptop production to software development, demonstrating the need to create specific nomenclatures when we look at any one of the forms of this work.
We use quotation marks in relation to this term because, formally speaking, such workers are not employees of the companies they work for.
For comparison, in 2010 the minimum wage was R\(510.00. With the dollar price closing at \)1.66 that year, 1.5 of the minimum wage in 2010 was the equivalent of $460.80.
This Brazilian economic crisis was related to the world economic crisis of 2008, which led to the failure of major banks, declines in profit rates and a worsening in the quality of life of the population, significantly increasing unemployment and eroding the purchasing power of the working class (Chesnais, 2013:22-5). The causes of the seven-year gap between these two crises is not one of the themes of analysis in this article, but we believe it is important to point out the existence of this world economic crisis of 2008, because some of the government policies adopted for its solution had a direct impact on ways of working.
In our analysis, both of these forms of labour management radicalise the real subsumption of workers to capital, with capital expanding its control over work processes and modifying them according to its interests, thus increasing the extraction of surplus value in its relative form.
On the subject of self-Taylorisation of work, especially among software developers, see Amorim & Grazia, 2018.
A feature of app companies is that they are registered as technology companies and not related to the services they provide, allowing the argument that their core activities are not outsourced, a practice that is prohibited in many countries. For example, in its terms of use (Uber, 2017), Uber do Brasil Tecnologia LTDA defines itself as a technology company responsible for maintaining a platform that allows users to request or provide transportation services, with this service being performed by independent third-party providers.
Highlighting how apps incorporate workers' know-how corroborates Braverman's (1981) thesis on capital's tendency to (re)qualify, disqualify and degrade work.
Returning to Srnicek's (2017) arguments once again, it becomes apparent that the huge amount of data retained by large platforms allows them to control a larger number of workers, and the more people networked generating data, the greater the value of the platform, a process that favours the creation of monopolies.
This percentage varies with each ride, it being the sole decision of the company how much will be retained. In our interviews, drivers reported that in some rides this rate could be as high as 50% of the total amount.
Uber allows different ways to pay for the ride in each country where the service is provided. In the Brazilian case, for example, Uber allows cash payment, which is not the case in other countries, due to the large percentage of the population without credit card access (Oliveira et al., 2019:20).
Although not becoming hegemonic in the capitalist mode of production, the piece wage was one of the principles of Taylorism (Taylor, 1990:88-9). Although the adoption of the conveyor belt has solved the problem of the pace of production in some industrial branches, the piece wage continues to be used as an instrument for increasing labour productivity. Some studies on the textile industry (Abreu & Sorj, 1993:51-2) and the sugarcane agribusiness (Tavares & Trindade de Lima, 2009: 173), for example, have demonstrated how this form of piece wage in Brazil is an important method for control of work and productivity, although often classified as a form of pay that can increase workers' freedom.
In order to reduce vehicle traffic during peak hours, the city of São Paulo has implemented, since 1996, the rotation of vehicles. Under this rotation system, each car cannot travel around the city centre one day a week, and with a scale organised by the final plate number (plates with endings 1 and 2 not being able to circulate on Mondays, 3 and 4 on Tuesdays and so on)
Hall and Krueger (2016) conducted a survey with Uber drivers in the USA and reported that 55% of respondents work as drivers for only 1 to 15 hours per week and that 60% of drivers do not have Uber as their primary source of income. The differences between these data and driver surveys in Brazil make it important to conduct a larger survey on the impacts of app companies on ways of working in a country in the global South.
An important feature of Uber drivers in the city of São Paulo is that 22% of these workers, according to the company, use a rented vehicle. The average rental price of a vehicle in São Paulo is R$ 400.00 per week, another factor that should be considered in the calculation of the net salaries of these workers.
We can say it is being consolidated by the workers just in time (Abílio, 2017; De Stefano, 2016; Oliveira, 2002:16). These workers are available for work 24 hours a day, rendering all existing time as potential working time, since remuneration is only for the hours actually worked. This makes the entire time frame a possibility for increasing incomes. In addition, the term just in time accurately describes how these workers are activated to perform their duties, being completely conditional on fluctuations in demand and the requirement to respond immediately to any work request.
The theory of human capital, developed primarily by the Chicago economist Theodore Schultz (1973), impels workers to continually invest in their acquired capacities in order to increase their market value.
Authors such as Woodcock & Johnson (2018) and Deterding et al. (2011) define as gamification some of the mechanisms performed by Uber to ensure the continuous provision of the service. Gamification would be the use of game design techniques in non-playful contexts to motivate the activity and loyalty of people in different activities. Through studies on the mechanics of gamers, platforms seek to develop procedures aimed at increasing user engagement in various platforms.
The minimum grade for the driver's disengagement from the platform varies according to the city in which the service is provided. In São Paulo the minimum grade allowed by the company is 4.65, and in other Brazilian cities this number ranges between 4.6 and 4.7.
One of the main demands of those attending the global strike of drivers by app in the city of São Paulo concerned safety. To illustrate the seriousness of the issue, data released by the Public Security Bureau show that in 2017 there were 3,952 cases of robberies of drivers while performing their activity. In the first quarter of 2018 this number was already 18.5% higher than the same period the previous year (Alcoverde & Perroni, 2018). A survey by drivers themselves shows that 55 workers have been killed in the state in the last three years. Importantly, in most of these cases the action was performed by passengers using fake accounts registered on the platforms.
Venco (2003:67) compares business software with the architectural figure of the panopticon (Foucault, 1999), as it serves to control all minor movements of workers, placing them in a conscious state of permanent visibility and automating hierarchical forms of power.
When we compare the work done by drivers by app and taxi drivers, we see that the functions performed bear several similarities. However, all this information quoted as being organised by the application in the work of Uber drivers is, in the case of taxi drivers, derived from their own experience as a worker, demonstrating how the adoption of algorithmic management acts to concretely modify the labour processes.