From Surveillance to Sousveillance: Designing Data Tools to Empower Platform-based Gig Workers 

By: Lindsey Schwartz, CDT Policy Fellow

Digital platform companies like Uber, TaskRabbit, and Amazon play a central role in today’s economy and our daily lives. In addition to connecting consumers to various products and services, these platforms connect workers with opportunities in a range of labor markets,   including transportation, domestic work, and knowledge work.  An estimated 16% of people in the United States have engaged in some form of platform-mediated gig work, with over half of those workers reporting that they rely on income from gig work for basic needs. 

Gig platforms currently have a near-monopoly on information about the markets in which they operate. Many platforms use their information advantages to increase their profit margins while making workers’ pay lower and less predictable. In an effort to remedy this information asymmetry, researchers and advocates have worked to develop tools for workers to collect and aggregate data about their working conditions, including sousveillance tools that (in contrast to surveillance tools) allow workers to collect information about the companies they work for. But the ‘gig’ economy is not a monolith, and the design and effectiveness of worker-driven data tools may differ for different types of gig work.

Engaging directly with workers from different gig economy sectors and platforms would help researchers and advocates design more effective technical and policy solutions to the problems created by the lack of transparency in gig labor markets.

Information Gaps in Gig Platform Labor Markets

By dint of their role as intermediaries for both commerce and labor, digital platforms collect and control vast amounts information about their users —– workers and consumers alike. This includes much information about workers and consumers that is sensitive (such as their identity and location) or of great economic value (such as the price a consumer is willing to pay or the wage a worker is willing to accept).

By collecting this information and controlling who can access it, gig platforms benefit from huge information advantages. Control of users’ information (again, workers and consumers alike) is a key part of many gig platforms’ business models. Platforms justify these practices by saying they help protect users’ sensitive information. But gig platforms also obscure critical information from users — such as how their wages are calculated — to manipulate workers’ behavior, often pushing them to work longer hours and endanger themselves in order to keep their performance ratings high. Information gaps can also allow bad actors to pose as customers or customer service representatives from a platform, making platform workers vulnerable to financial scams, wage theft, and assault.

Workers and regulators alike often do not know the extent of the data that platforms collect and retain from workers, including how platforms use this data to refine the algorithms that shape working conditions. Platforms’ proprietary treatment of their data practices leaves policymakers and advocates with limited information on the gig workforce and working conditions, making it difficult to design, implement, and enforce laws to protect gig workers. For workers, the lack of transparency further weakens their capacity for collective action and advocacy — which is already limited due to their isolation from one another in the fissured workplace. The upshot of the platforms’ information advantages is that neither workers nor regulators can hold them accountable for their harmful practices. 

Prior Research

To combat gig platforms’ information advantage, researchers have been designing and building digital tools that enable workers to independently collect and analyze data about the gig markets in which they work. This approach draws from the concept of sousveillance. Sousveillance is the inverse of surveillance — instead of the platform monitoring its workers, sousveillance allows workers to monitor the platform and, in some cases, the clients with whom the platform connects them. These tools allow workers to overcome information gaps and better navigate the contingencies and risks they encounter on the job, with the goal of empowering workers both individually and collectively. Examples of sousveillance tools include the Shipt Calculator, which allowed Shipt’s delivery workers to audit the platform’s price-setting algorithm; Turkopticon, a tool for workers on Amazon Mechanical Turk to rate clients; and GigSense, a collaborative troubleshooting tool for workers on Upwork. 

Researchers have identified a number of factors that affect the sustainability of these initiatives, such as the resources required for long-term maintenance of digital infrastructure and the risk of retaliation against workers who use these tools. Researchers should also take into account another fundamental factor: the widely different contexts within which platform workers work.

Eurofound Typology of Gig Labor Platforms 

Eurofound, the labor agency for the European Union, published a typology in 2018 that categorizes platforms based on different combinations of the location of service provision (on-location or online) and the selector (which party determines where and how the work takes place).

In addition to the key features identified by Eurofound, additional criteria may help determine how workers can benefit from sousveillance and other data tools, such as the degree to which workers communicate and negotiate directly with clients, how pay is determined, and how many parties are involved in the transaction. For now, however, the Eurofound typology provides a helpful framework, and we use it to illustrate how a few of the key differences of gig work may be relevant to the design of sousveillance tools.

Key differences in types of gig platforms

In platform-determined work, the platform largely controls the client-worker matching process and can adjust prices and wages based on supply and demand. Algorithmic management plays a significant role in shaping working conditions on these platforms. Instead of being shown all available tasks, workers receive select offers through the app based on factors such as location or performance ratings. While they are given the option to reject offers, they are deterred from doing so because of the risk doing so may drive down their performance ratings. 

Platform-determined work can take place on-location (e.g. ridehail driving on Uber; food and grocery delivery on DoorDash), where workers must travel to specific locations assigned by the platform, or online (such as crowdwork on Amazon Mechanical Turk), where workers may be located anywhere around the world.

In client-determined work, the client — which may be an individual or a company — specifies the task that is to be performed. Workers may perform the task on-location (e.g., care work, hospitality work, and ‘handyman’ repair services) or online (such as coding,  web design, and digitally-mediated forms of sex work on platforms such as OnlyFans). In these arrangements, workers are not algorithmically matched with clients, but instead have to indirectly compete for opportunities and maintain high performance ratings for credibility. Compared with workers in platform-determined work, workers in these client-determined platforms may experience a higher degree of autonomy in how the job is completed, as well as the ability to take the worker-client relationship off the platform.

How these differences might affect sousveillance tool design

While information gaps are a shared feature across these platforms, the differences between these types of ‘gig’ arrangements may translate to a range of sousveillance tools tailored towards specific issues and worker characteristics found in each setting. Workers in platform-determined work often advocate for increased transparency around aspects of their working environment, such as platforms’ calculation of wages and customer verification, due to concerns about their finances and safety. Potential interventions in client-determined work may have similar aims but be instead directed towards improving client-worker relations, since the client rather than the platform is supervising the work. For example, workers could benefit from tools that help hold clients to the same credibility standards as workers, such as a tool to rate and share feedback about clients with one another. 

Location (online versus on-location) influences differing definitions of workplace safety and, therefore, the potential aims of sousveillance tools. Workers who are based online (eg. knowledge workers) are often concerned with digital privacy, while on-location workers — such as ridehail drivers — are concerned with physical safety in addition to digital privacy. Workers in both settings may benefit from tools that enable them to collectively identify and troubleshoot shared issues, while video recording tools are most useful for on-location workers. 

The need for worker input and research

In order to account for the different contexts within the gig economy, researchers and advocates should seek input from workers across these categories when assessing the potential for sousveillance tools. Such engagement should focus on determining:

• What kinds of information are valuable to workers
• What strategies they currently utilize to mitigate the information gap in their workplace,  including what kinds of data they are able to collect
• What questions they have about their work environment
• What concerns they have about using sousveillance tools 

The answers to these questions will give researchers and advocates a clearer sense of the issues gig workers face and what kinds of information would help workers investigate the contingencies of their workplace and increase their individual and collective power. Crucially, this approach would enable researchers to bolster existing advocacy efforts led by workers, whose strengths and networks tend to be undervalued.

Sousveillance tools have great potential to mitigate the information gaps in the ‘gig’ economy by supporting workers in their individual and collective efforts to improve their lives, livelihoods, and working conditions. Aligning the design of these tools with workers’ preferences and priorities can lead to tangible improvements to working conditions across platforms, as well as contribute to our broader understanding of platform labor and the platform economy.