Towards Fairer Work in the Digital Gig Economy

When we use a product, a service, or even an algorithm that was brought into being with digital labour, there is no way to know whether an exhausted worker is behind it; whether they get laid off if they become sick or get pregnant; whether they are spending twenty hours a week just searching for work; how precarious their source of income is; or whether they are being paid an unfairly low wage. This proposal envisions a way of holding platforms in the gig economy more accountable through a programme of research focused on fair work, a ‘Fairwork certification scheme’, and a ‘Fairwork platform ranking.’ It operates under a governing belief that core transparent production networks can lead to better working conditions for digital workers around the world.

According to the World Bank, there are now 48 million digital workers that live all over the world, doing work that is outsourced via platforms or apps in the gig economy. Lacking the ability to collectively bargain, workers have little ability to negotiate wages and working conditions with their employers — who are often on the other side of the world. Our previous research has shown that, as a result of this global market for work that has been brought into being, many workers have jobs characterized by long and irregular hours, intense work, low income, and high stress. The international nature of much digitally mediated work means that work also tends largely to be done outside of the purview of national governments, with very few employers paying attention to relevant regulation on the books in either their home countries or the worker’s home country. Previous research has shown that international digital labour platforms thus threaten to undermine workers’ ability to defend existing jobs, livable wages, and dignified working conditions, in both low- and high-income countries.

This is a state of affairs that is not just undesirable for workers, but also for client firms and end-consumers. Client firms will want to avoid the reputational risks of outsourcing to poorly-treated workers; and research has shown that consumers who are able to do so are often willing to pay a premium to ensure that products they buy were produced under good working conditions.

We thus propose a comparative programme of research that uses a mixed methods approach (combining quantitative and qualitative, as well as digital and offline, methods) in order to assess working conditions in three types of gig economy platforms:

• crowdwork platforms: platforms for highly commodified work where clients never interact directly with workers; e.g., Amazon Mechanical Turk or Crowdflower
• freelance platforms: platforms that facilitate a more direct relationship between client and worker; e.g. Upwork or Freelancer
• location-based apps: apps requiring workers to be in a specific location; e.g. Deliveroo, TaskRabbit, or Uber

Certifying and Ranking Fair Work

This programme of work aims to not just uncover where fair and unfair work takes place, but also seeks to codify that knowledge into both a ‘Fairwork certification scheme’ and an annual ranking of platforms. These two initiatives build on the efforts of organisations such as the Fairtrade Foundation and will ultimately allow for the development of an international standard for good-quality digital working conditions.

This project will draw on the expertise and experience of Oxford Internet Institute staff on work practices and working conditions in digital labour platforms. Project staff will further refine criteria for fair digital work under development in collaboration with researchers at the International Labour Organisation, Eurofound, the German Ministry of Labour, IG Metall, and organisations with expertise on working conditions generally. Specifically, we will develop a set of criteria based on nuanced empirical understanding of work processes within platforms. The criteria will take note of the differences between the three platform types and will be divided into two levels. That is, after assessment, a platform may receive no certification, or may receive a certification of acceptable or good working conditions. To receive an acceptable certification, platforms must meet the following criteria:

1. Minimum wage. Platform ensures that workers are paid at least minimum wage in their location, regardless of employment classification.
2. Regulation of nonpayment.
   1. Freelance and location-based platforms: The platform does not allow non-payment for completed work.
   2. Crowd work platforms: Either the platform does not allow non-payment for completed work, or nonpayment is strictly regulated. Specifically:
      1. Customers who refuse payment for work are required to indicate in a legally binding manner that they will not use it.
      2. Customers must explain why the work was unusable.
      3. Workers have a right to contest nonpayment. If a worker contests, the claim is reviewed by a human employee of the platform.
      4. Customer nonpayment rates should be made visible to workers when choosing tasks.
      5. Task-specific conditions under which nonpayment will be permitted are specified in advance and are visible to the worker before they accept the task.
3. Compliance with relevant laws. The platform abides by all relevant laws in the worker’s location (e.g., employment classification, occupational health and safety, intellectual property, contract law, right to organise).
4. Pay terms.
   1. For crowd work and freelance platforms, the time in which clients agree to review and pay for submitted work is stated up front, and is clear to the worker before accepting the task.
   2. For crowd work platforms in which nonpayment is permitted, task-specific conditions under which nonpayment is permitted are clearly specified.
5. Non-competition agreements. The platform does not require workers to sign non-competition agreements.
6. Non-disclosure agreements. If the platform requires workers to sign non-disclosure agreements, the agreement prohibits disclosure only of data submitted by customers, not pay, work processes, or working conditions.
7. Access to collected data. The platform allows each worker access to all data collected about them by the platform at any time, including work history data and work evaluations or ratings. That is, the platform allows each worker to see the “file” compiled by them at any time.
8. Contestation of work evaluations or qualifications. The platform allows workers to contest work evaluations and qualification test outcomes. Such contestations are reviewed by a human employee of the platform.
9. Communication. The platform ensures that customers and platform operators respond promptly, respectfully, and substantively to work-related worker communications.
10. Information about client and purpose of work. The platform gives workers information about the client and the purpose of their work.
11. Psychologically stressful or damaging tasks.
    1. Tasks that may be psychologically stressful or damaging (e.g., review of social media content for hate speech, violence, or pornography) are clearly marked.
    2. Workers completing such tasks have access to counseling or support paid for by the customer and/or platform.
12. Right to collective representation and bargaining.Regardless of employment status, workers have a legally protected way of collectively negotiating with platform management for changes to work processes and improvements in working conditions — e.g., through trade union representation and collective bargaining.
13. Account deactivation. Worker account deactivations are reviewed by a human platform employee. Workers may contest deactivation and have the contestation reviewed by a human platform employee.
14. Pre-publication review of task instructions. For crowd work and freelance platforms and for non-standardized tasks in location-based platforms about which workers are unlikely to communicate directly with clients, task instructions are reviewed by a human platform employee or a qualified crowd worker prior to publication.

Data Collection Methods

We propose a three-part data collection strategy for assessing work processes and working conditions on the platforms:

1. Collection of public data
2. Surveys and interviews with workers
3. Digital and offline participant observation

Outcomes

There are three key outcomes for the project. First, the research will provide much needed insights into the working processes and practices on digital labour platforms. Second, and building on the efforts of the ‘ethical trade’ movement, the dissemination of the project findings will highlight working conditions on the selected platforms, along with indicative rankings and certifications in each area. Third, the longer term aim of the project is to encourage platforms to evolve their processes and practices in ways that improve working conditions.

Many of the millions of digital workers who carry out gig work have very little bargaining power. Their ability to collectively bargain is limited, and they are often not protected by existing rules and regulations. As ever more people come online looking for jobs, the prospects for workers collaborating instead of competing look bleak. A programme of research and action focused on fair work offers viable strategies to change that by building a knowledge base about wages and working conditions in the gig economy, and encouraging employers, consumers, and platforms to improve them.

Contact

Mark Graham, Oxford Internet Institute
Jamie Woodcock, London School of Economics
Ellie Harmon, Encountering Tech, LLC
Six Silberman, IG Metall (German Metalworkers’ Union)