Being able to order a rental vehicle with driver (VTC) via an application seems to revolutionize the market, which has until now been regulated in the taxi sector. Once established in large cities, these new platforms leave behind a trail of suspicion about the functioning of their algorithms. Several studies have shown that these types of companies take into account the battery condition and the model of the mobile phone from which the service is provided to set prices. The administration, always slower than technology, tries to make public the variables on which the final price is calculated. However, experts warn that this may not be enough.
“Algorithms are presented to us as if they were something neutral that discriminates only based on supply and demand, but that is a lie. “Algorithms respond to the interests of the companies that create them, whose goal is to achieve maximum profit at all costs. » This is explained by Tito Álvarez, spokesperson for Taxi Project, an organization that has studied in depth the darkest points of the system on which platforms like Uber, Cabify and Bolt rely.
Álvarez and Sergi Cutillas commissioned a report in which they analyze Uber’s processing of data. “The algorithm used to calculate travel prices is one of the company’s most protected assets,” the document introduces. Despite this, and although the data used by the algorithm and how it is used is unknown, we see that the end result is the regulation of supply and demand by changing the price, they say.
To reach some conclusions, Taxi Project experimented with how the algorithm responded to requests for similar Uber rides, simply by varying the mobile phone model and battery percentage. The variations were small, they point out, “but each time they occurred, the price was higher for the device with a weaker battery, especially when the course ended in a stadium.” “They do this to play on the customer’s desperation,” explains Álvarez. Additionally, they checked how the price increased if the service was requested from a high-end mobile phone.
Possible “algorithmic cartel”
The awareness that Taxi Project is trying to make society aware of the dangers of using these technologies could not take effect without the computer scientist and hacker they have on their payroll. “Through their research, we saw the correlation between Uber and other ride-hailing services. It seems that when Uber raises its prices, so do Cabify and Bolt. We do not know if this is a kind of price pact between algorithms or if the companies use the same indicators to increase and decrease them,” explains the spokesperson for this group.
For its part, Taxi Project has already informed the National Market and Competition Commission of its investigations into a possible “algorithm cartel”. Álvarez, as a taxi professional, adds that the most worrying thing in this matter is that these companies belong to investment funds that “determine prices to eliminate competition and, once obtained, increase them without any form of regulation “.
Uber, an extrapolated system
Víctor Riesgo is a researcher in the Critical Urban Studies group associated with UNED and his thesis was entitled Control, consent and resistance in platform work. Driving for Uber in Spain. From their perspective, the algorithm is capable of discriminating workers based on hidden variables. In this sense, the consultancy firm Games Econ has determined that during the Mad Cool festival in Madrid in 2023, the prices of these companies increased by up to 250%.
This expert sees beyond the cost of travel. “We use these apps without knowing very well what permissions we are allowing. They know our shopping habits, where we travel, our schedules, even what we eat, if we use Uber Eats,” he illustrates. Anyway, for Riesgo there is another indicator more worrying than that of the final consumer: “In the case of workers, everything is more complex and complicated because they are obliged to reach certain billing levels, in addition due to the fact that they often use their personal cell phones for work.
In the eyes of this researcher, the concept of “Uberization of society” would not be very far from reality. According to him, “Uber’s aspiration, I believe, is to become a kind of bits that can be managed and processed from binary codes.” The company would thus carry out “a long-term personnel management project, since its system can be used in other departments to assess the workload, distribute it and assign a price”.
Workers, unaware of the profits of the algorithm
On the same side of the screen are the drivers of these large companies, which the platforms also use to make money. “We don’t interfere with the algorithm at all, but I feel like there must be some sort of price agreement between them, because when there is high demand at Uber, there also has strong demand in the rest”, explains Juan Fernández, Uber. driver and union representative of CC OO of the Moove Car group in Madrid, one of the three companies taking over VTC licenses in Spain, managing up to 85% of them in the capital through employees.
“I don’t work with incentives, so as a worker I don’t worry about price increases. This goes to my company and to Uber, which earns around 25% for each service provided,” emphasizes Fernández. The Moove Car works council asked to know the algorithms, but the answer was negative. “They tell us that it does not depend on their company, that it belongs to Uber and that they do not have access to it, which is yet another way of blocking the demands of legal representation of workers,” affirms this trade unionist.
Companies defend themselves
Uber assures that its price “is defined according to the combination of the base price, the duration and the distance of the trip”. “The algorithm does not discriminate in any way by taking into account factors such as the battery of the mobile phone from which the service is requested or the model of the device,” it adds. In addition, they defended these “dynamic prices” to increase the supply of drivers, for example during large events, and “ensure that users have access to a vehicle”, in their words.
Asked about this question, Cabify assures that its technology “calculates the price of a trip based on time (minutes) and distance (kilometers)”, taking into account “concepts duly published and available to users on their page web”. and in the application. They also deny that the company takes into account factors such as battery or phone model when establishing the price of a trip, nor any variables related to personal, social or other aspects.
From Bolt they answer something similar: “Bolt’s dynamic prices depend mainly on circumstantial factors such as location, length of trip and market conditions at the time of booking,” they explain. Additionally, they deny that their system takes into account “the specific device or device manufacturer the order is placed with, nor data related to past customer behavior or phone model battery.”
While this company works with more than 8,000 VTCs and taxis in Spain, Cabify preferred not to provide this information, while Uber did the same “for competitive reasons”. None of the three companies responded on the percentage they earn for each trip taken.
Make algorithms public
Sumar registered an initiative in Congress to regulate these algorithms. They want the platforms to have “a published and updated description, available to the user and in a manner sufficiently understandable for the citizen”, in the words of MP Tesh Sidi, of the variables which determine their functioning . On the other hand, they ask the CNMC that these companies cannot determine their prices based on aspects such as nationality, place of residence, place of establishment or request for service or the type, age or the charge level of the electronic device used.
From Taxi Project they welcome the idea of making public the variables on which the algorithms of this type of platforms work. “They say it’s a business secret, but in reality it’s a consumer right,” defends Álvarez. In any case, the group’s spokesperson is not very confident about the realization of this measure: “They will fight tooth and nail because there is the mother of the lamb, their whole business.”
Riesgo, for his part, considers it somewhat “naive” that the publication of the algorithms, which Sumar does not even require, could substantially change reality. “These companies’ machine learning models are so complex that it’s very difficult to be able to interpret the variables,” he explains.
This expert advocates the creation of a public platform for this type of transport. “The idea is that the state creates a platform from which these companies could provide their services according to common standards and rules, even if this would go against the neoliberal logic in which we live, where the train has been liberalized and some bus lines also sell their tickets based on supply and demand,” he concludes.
The Administration, behind
Álvarez, spokesperson for Taxi Project, says that “the technology is well ahead of the Administration’s regulations.” Furthermore, every time they met with public bodies to denounce the system by which these platforms operate, they found themselves with a feeling of strangeness. “They look at us like we’re aliens. They don’t see that we are moving towards a hyper-technological society. In the times we live in, unions should have more computer scientists and fewer lawyers,” he comments.
Riesgo, a researcher at UNED, considers that administrations are “following behind, without any initiative and, unfortunately, following the interests of the large platforms”. Even if some progress has been made in terms of transparency at the European level, he believes that little by little “private companies are being granted the capacity to regulate social and public life, as is the case in “X” for this who is discussed in each moment and from what position this is done,” he gives an example.
