Can workers access their data and use it to build their power? The Gig Economy Project looks at a fascinating case study from Geneva, Switzerland, to find out.
Every worker in Europe has the right to request what data is held on them by the companies they work for under EU GDPR rights. The mechanism for doing this is called a Subject Access Request (SAR). One worker who has tried to use this method is Guillaume [full name withheld for anonymity purposes], an Uber driver in Geneva, Switzerland, from 2017 to May 2022.
Guillaume has had the support of PersonalData.io and HestiaLabs in trying to make sense of the data he has been able to attain from Uber. As will soon become evident, this is no easy task.
This data pursuit also comes in the context of a major dispute between Uber and its drivers in Geneva (including Guillaume) over how much back-dated pay they should be owed, after a court found that the drivers were employees. How much money the drivers are owed can only be accurately assessed via their data.
Can workers access their data and use what they uncover as the basis for trade union demands, legal cases and political action? The Gig Economy Project (GEP) has taken a deep dive into Guillaume’s data discovery to find out.
What does Guillaume’s data show us?
“I am a little bit of a geek, I like games on the computer, and I’m interested in data,” Guillaume tells GEP, in explaining his motivation for trying to access his data at Uber.
Guillaume had begun recording his data in real-time from his app dashboard while working, and then sought to compare that to what he got paid to see if he was being cheated – a very laborious process. He was therefore pleased to come across a potentially more efficient way to do this when he saw a PersonalData.io video on social media about how to recover your data.
The subsequent collaboration has went well beyond analysing Uber’s SAR response to Guillaume, also taking in an Uber download portal file, information on uber.drivers.com, screenshots Guillaume has taken of the app while at work, and data from other geolocation apps (such as Waze). The analysis of this work has been collated in a “technical audit”. What does this data trove tell us?
First, that it is “almost impossible” for a driver to get any meaningful information from the data Uber provides without the help of data specialists. The data’s “formatting and presentation is more useful to an internal Uber employee than a driver,” the audit states. Problems include:
- The data was only given in English, and used miles rather than kilometres (the standard measurement in Switzerland)
- It was badly organised: there are no clear definitions for each data type, meaning the driver has to try to work it out themselves, and data was given in different timezones
- There was accuracy issues, for example the same trip had significant discrepancies across files received, including the incorrect calculation of fares
- The data is incomplete: for example, the fares refer to the total paid by the client, not how much the driver earned after Uber had taken its commission. There was also less data given than what the driver can even see on their live dashboard on the app, geolocation data was only given for a two month period, and app connection data for just two days
The inadequacy of the response has been of little surprise to Guillaume.
“The process is so difficult to contact Uber and get something from them,” he says.
“I compare the data from the screenshots and I think the data that they gave me is not right,” he adds.
Jessica Pidoux, director of PersonalData.Io, says Uber’s response to Guillaume’s SAR “could be a breach” of GDPR rights. For instance, Article 20 of GDPR is “the right to data portability”, meaning “the data subject shall have the right to receive the personal data…in a structured, commonly used and machine-readable format”.
“it is definitely not in such a format today,” Pidoux argues.
Secondly, for any data not to be included under GDPR, the company must justify the risk that it would pose to provide this data in a risk assessment, something Uber did not do in its response to Guillaume.
“For example, they gave Guillaume two months of geolocation data, and they answer that ‘we cannot give away so much you have to give a more specific time period as you can put others at risk’.
“We consulted a data protection lawyer, who said that Uber have to prove how they would be putting others at risk by including this data. And the other thing is you can anonymise the data, there are techniques that you can apply, but there has to be a willingness from the company to do this.”
The technical audit set out to find answers to three simple questions about Guillaume’s data:
1) ’What distance have I driven in a given time period?’
2) ‘How does my revenue in a given time period compare to the amount earned by Uber?’
3) ‘How much time do I spend waiting in between rides while working?’
Without the help of personaldata.io and HestiaLabs, Guillaume would have not been able to find answers to any of these questions from the data files he received from Uber. But after the combined efforts of Guillaume and the data experts, they were able to find some, though by no means all, of the answers.
1) Various figures for the distance driven result from the analysis, “leaving the driver in a position where they are unable to say with confidence which is the accurate calculation of the distance driven in a given time,” the report states. The most accurate figure appears to be between 579 and 582 kilometres for a week in July 2019, since these are the figures of both the data from Uber’s download portal and the screenshots from the app.
2) Again, there are inconsistencies in the figures for the revenue Guillaume received, and it’s not possible to get this information from the SAR response at all. But based on the information in the data portal and the app screenshots, a figure of 1369.61 Swiss francs can be arrived at.
3) A huge amount of work was required on the raw data to come to an assessment of waiting time, including calculating the average time Guillaume had the app open but was working for other private hire platforms. The research shows that Guillaume’s waiting time increases dramatically from 2020 onwards, rising from 35.4% in 2019 to 81% in 2020 (see table below).
According to Guillaume, two things changed in 2020: the pandemic “caused the number of orders to drop” and in July 2020 Uber introduced a new multiplier system for drivers whereby they can raise or decrease the fare they will receive for an order, but this will affect their chances of getting the order.
“My rate was often too high for the algorithm, and so I received very few orders,” he says.
The fact that Guillaume’s waiting time more than doubled in one year gives some indication as to the extent to which algorithmic changes can affect the experience of being an Uber driver.
Guillaume’s trips and waiting time data, 2017-2022
| En route | On trip | Waiting time | Waiting time % | |
| 2017 | 22:52:01 | 1 day, 19:30:30 | 1 day, 5:38:58 | 31.3 |
| 2018 | 4 days, 13:12:52 | 9 days, 23:43:40 | 4 days, 11:54:23 | 23.6 |
| 2019 | 11 days, 02:56:17 | 27 days, 00:31:51 | 20 days, 21:15:20 | 35.4 |
| 2020 | 4 days, 15:15:20 | 9 days, 20:45:58 | 61 days, 17:03.45 | 81 |
| 2021 | 2 days, 03:52:51 | 4 days, 04:09:29 | 55 days, 13:44:51 | 89.8 |
| 2022 | 11:58:22 | 14:17:12 | 8 days, 21:43 | 89.2 |
What else can be deduced from this data about Guillaume’s work? Well, it’s possible to calculate the difference between the total fares taken and Guillaume’s revenue, meaning we can work out Uber’s commission over one week in July 2019 (448.98 Swiss francs, or 24.1%). These figures could be useful, for instance, to make a pay claim.
Finally, there are other issues highlighted by the technical audit which are worth highlighting. One is that Uber continued to collect data on Guillaume after May 2022, when he stopped driving for the company. It is therefore clear that Guillaume is “tracked whether he is working or not, whether he opens the app or not,” the report states.
Secondly, the one file which mentions explicitly that Uber has passed on Guillaume’s data to an advertiser refers to Californian users’ rights, seemingly ignorant (or uncaring) of the fact that Guillaume is based in Switzerland. This “demonstrates that Uber does not make an effort to comply in any way that is meaningful to users not based in the United States,” the report concludes.
Geneva drivers’ battle with Uber
After a court found in May that Uber drivers in the canton of Geneva were employees, and thus were entitled to salary arrears from the US company, there has been a battle over how much back-dated pay the drivers are owed.
The company offered 4.6 million Swiss Francs to cover back-dated pay from 2017 to June 2022, but that was rejected by unions, whose lawyers are demanding access to each individual drivers data so that they can work out what the back-dated pay should be. Suddenly, the work of personaldata.io and Guillaume has become relevant in a very direct way.
Previously Guillaume and Pidoux had struggled to get the workers’ lawyers to see the importance of the data, but they are “asking for the data because they now realise it is important, but they don’t know how to read the data,” Pidoux says.
“They only have work lawyers in there, and what we are pushing with them is that they need a data protection lawyer. We have been offering our expertise for years, but the unions have always said ‘it is not the moment, it’s not important’.”
Pidoux adds that it’s not enough to analyse the data which Uber provides the lawyers, the drivers need to make their own individual submissions for their data so they have something to compare against what Uber provides.
“If drivers gather together to get their data it’s going to be the only way they build more power and visibility on Uber,” she argues. “It’s about bringing forces together to respond collectively to Uber; for us it’s the only way to do a big change.”
It took five workers plus Guillaume to analyse his data, including an engineer, a data journalist and a mathematician. Pidoux says organisations with much bigger resources than PersonalData.IO and Hestia Labs need to start dedicating resources to data work.
“That’s why unions need data experts, lawyers need data experts, and the state needs data experts,” she argues. “The idea of data transparency is very beautiful, but who is going to do it?
“We are data experts but we want that society has more experts on data, and the problem is organisations don’t have the finances or they don’t see the relevance, but it’s coming.”
Of course, data analysis would be made much easier if regulations were tightened so that companies like Uber had to respond to a SAR with data that is legible, accurate and complete. Guillaume is hopeful that a new Swiss law (which entered into force on 1 November) to regulate private hire platforms will increase the data obligations on companies, and says he will only re-enter the platform economy if data transparency becomes a reality.
“We have to change the laws,” he says. “I want more transparency of the Uber algorithm, because we don’t know how Uber works, and I want to know what is happening inside the app. If I don’t know, I don’t want the work. I like to understand what I do.”
Pidoux says there are powerful forces at work to prevent easy access to data of the giants of the platform economy.
“Firstly, I think the regulators don’t understand data, or else they could require companies to provide data structures, very specific things,” she says.
“Secondly, they don’t want to block the economy, and platforms are giving a lot of money now to the state in taxes. Then you have intellectual property, trade secrets, etc. What it amounts to is that the platforms are protected.”
For workers to build data power, the battle will have to be fought economically, legally and politically. But the starting point is to understand that data is power, and therefore for workers not to have access to and analysis of their data is a limiting factor on their power. Guillaume and PersonalData.Io, along with the likes of Worker Info Exchange in the UK and the INV union in France, are pioneering a path forward.
To sign up to the Gig Economy Project’s weekly newsletter, which provides up-to-date analysis and reports on everything that’s happening in the gig economy in Europe, send your e-mail to GEP@BraveNewEurope.com.
Support us and become part of a media that takes responsibility for society
BRAVE NEW EUROPE is a not-for-profit educational platform for economics, politics, and climate change that brings authors at the cutting edge of progressive thought together with activists and others with articles like this. If you would like to support our work and want to see more writing free of state or corporate media bias and free of charge. To maintain the impetus and impartiality we need fresh funds every month. Three hundred donors, giving £5 or 5 euros a month would bring us close to £1,500 monthly, which is enough to keep us ticking over.
Be the first to comment