Digitaler Turmbau zu Babel
Gesine Born/KI: Midjourney

Not fair at all

Working conditions on online platforms for language services

It's hard to imagine working life without the language services provided by digital platforms. However, it raises the question: How fair are the working conditions for employees engaged with these platforms? To shed light on this matter, a team of researchers from the Fairwork project conducted an in-depth examination of nine platforms, assessing them based on five principles of fair work.

The digital platform economy has been seeing steep growth. The same goes for the market for language services that can be performed remotely. Outsourcing of services such as translation, transcription, or subtitling is already common in a wide range of sectors: transcribers working on online platforms help law firms transcribe court proceedings, turn recordings from medical consultations into written records, document emergency calls for local authorities, generate minutes for business meetings, or transcribe interviews for researchers. Demand for translation services is also increasing. Global businesses need to offer their websites in multiple languages to address customers in various markets. Similarly, multinational, globally distributed teams require processes and documents in multiple languages.

Digital translation and transcription labor platforms intermediate these services. By pooling workers with different language skills from a variety of countries, they provide clients with a simple one-stop shop. The convenient and cheap services offered on digital labor platforms are appreciated worldwide.

However, the impact of this new form of gig work on workers is far more difficult to pin down. Freelancing work arrangements have been common in the translation and transcription sector before the advent of digital labor platforms. Now, these platforms are opening up new job opportunities and earning potential for many people, especially workers in low- and middle-income countries. The emergence of digital labor platforms is thus intensifying global competition between workers and shifting the balance of power in favor of clients.

But who defines how and under which conditions clients and workers can access this market? As a rule, it is the platforms that determine business models and practices. They largely define the terms of exchange unilaterally and often use non-transparent algorithms to set prices or manage work allocation. This power imbalance has consequences: The platforms’ different business models and practices determine the working conditions of millions of individuals providing online translations and transcriptions. Conversely, however, this also means that positive changes in platform practices can quickly improve conditions for many workers.

This is the starting point for the Fairwork Project, a collaboration between the Oxford Internet Institute and researchers at the WZB. It investigates and evaluates working conditions on various digital labor platforms. On the one hand, the research project contributes to a better understanding of the transformation that digital labor platforms are causing in the way we work today. On the other hand, the project aims to make an active contribution to a fairer platform economy by highlighting differences in the platforms’ business models and practices, which have significant effects on the working conditions of platform workers. Consumers are thus given the opportunity to make informed choices when selecting platforms. This also provides platforms with an incentive to actively improve their working conditions. The Fairwork Project is always happy to serve as a point of contact for those platforms that agree to implement improvements on their own platform, and to provide advice on industry-wide good practices and experiences.

Today, the Fairwork Project produces rankings of place-based platforms in 38 countries and of cloudwork platforms not restricted to specific labor markets.

The Fairwork Translation and Transcription Platform Report was published in October 2022. It evaluates nine platforms according to five principles of fair work: fair pay, fair conditions, fair contracts, fair management, and fair representation. For this purpose, the Fairwork Project assessed publicly available information and conducted interviews with workers and platform management. For each principle – each having two thresholds – a maximum of two points could be achieved by each platform, if all criteria for both thresholds were met (for details on the principles and the related thresholds, see here: Link). Only platforms that were awarded the point for the first threshold of a principle (e.g., 1.1) were able to receive the point for the second threshold (e.g., 1.2).

As the ranking shows, the platforms that were scored remain far from achieving the basic standards of fair work. No platform scored all ten points. However, the ranking also highlights the vast differences in working conditions on the platforms included.

Four platforms (Translated, TranscribeMe, Gengo, Lionbridge) received one point for having systems in place to ensure that workers are paid in a timely manner for all the work they completed. Of these four platforms, only one platform (Translated) was awarded the additional point for showing that the vast majority of active translators on the platform earn at least their local minimum wage.

Fair conditions

Three platforms (TranscribeMe, Translated, Lionbridge) were awarded one point because the Fairwork researchers found evidence that these platforms take steps to actively manage the supply and demand of labor on the platform, such as restricting sign-ups of new workers, thereby reducing competition between workers for available jobs. Of these platforms, Translated and TranscribeMe received an additional point for demonstrating active measures to protect workers from work-related health and safety risks, such as psychological stress from exposure to explicit or disturbing content in audio files or text documents.

Fair contracts

Only two platforms, Translated and TranscribeMe, were awarded points for fair contracts. For both platforms, the Fairwork team found evidence that contracts are written in clear language and accessible to workers, consistent with the platform’s terms of engagement on the platform, and that contracts do not contain clauses requiring workers to waive their right to legal recourse.

Fair management

Three platforms (TranscribeMe, Translated, Scribie) met all the criteria for fair management.  The research team found evidence that workers can communicate with a human platform representative. Moreover, there are officially documented and effective processes for workers to appeal decisions such as bad reviews or disciplinary actions. In addition, researchers found evidence that workers receive explanations for all punitive actions. TranscribeMe and Translated received an additional point for demonstrating the existence of a policy to mitigate the risk of discrimination against workers by the platform or clients.

Fair representation

The study was unable to award a point for fair representation to any of the researched platforms due to a lack of evidence of dispute resolution processes in which workers have access to an independent advocate or documented procedures for collective dialogue and bargaining.

Overall, the study highlights a need for action to improve working conditions in this sector. Platforms can offer low prices to their clients by avoiding any fixed costs for labor and instead sourcing labor power from “on-demand” labor pools. Since workers on cloudwork platforms are usually freelancers, they are usually not covered by national labor laws and thus are not protected by basic rights and benefits such as minimum wages or paid sick or parental leave. As a result, cloudwork in the translation and transcription sector can provide new, attractive income opportunities, especially for workers from low-income countries, but in most cases work remains precarious and insecure.

However, the study also highlights differences between platforms, demonstrating that platform work need not be unfair. Consumers of translation and transcription platforms are thus given the opportunity to help improve the platform economy through their choices.

For better working conditions on online platforms: the “Fairwork Pledge”

The Fairwork project specifically calls on organizations that make use of digital platform labor in their business operations. Organizations are invited to use the Fairwork ratings as a criterion in their choices and to make this commitment public by supporting the Fairwork Pledge.

The pledge constitutes two levels. At the first level, an organization can become a Fairwork Supporter by publicly declaring support for fair labor practices on platforms and making resources available to staff and members with resources to help them in deciding which platforms to engage with.

At the second level, organizations can become official Fairwork Partners by voluntarily committing to concrete and binding changes in their own processes and practices, for example by using better-rated platforms in their procurement, a policy pioneered by the WZB in 2021.

More about the Fairwork Pledge

Literature

Graham, Mark/Anwar, Mohammed Amir: „The Global Gig Economy: Towards a Planetary Labour Market?“ In: First Monday, 2019, Jg. 24, H. 4-1. DOI: 10.5210/fm.v24i4.9913.

Fairwork (2022): Translation & Transcriptioon Platform Ratings

Image description: Software surrounds and permeates working life in software-based organizations. Our photo editor Gesine Born used the Midjourney software to create this image using artificial intelligence. She entered the following keywords: "a photo of The Tower of Babel with modern lights, poor hard working people with laptops in a slum in front, blue sky, sunlight, --ar 4:2 --v 5.1 --s 50".

This text is licensed under a Creative Commons Attribution 4.0 International License.

17/7/2023

Download this article as a PDF here.