Content moderators and Post Traumatic Stress Disorder with Janine Berg

Janine Berg is an expert on digital labour platforms: she has been studying them since 2015, as part of her larger work on non-standard employment.

She is also an economist with the International Labour Office, the Secretariat of the International Labour Organization. Here research focuses on transformations in the field of work and the regulation of labour markets and provides technical assistance to ILO constituents on policies in support of decent work.

Her responses to the questions are her opinion and not the official view of the ILO.

How would you describe microtasking and how important is it for our economy and the economy of big tech?

Microtask platforms are crowdwork platforms that provide businesses with access to a large flexible workforce for the completion of small, often repetitive, clerical tasks.

Leveraging the power of “the crowd”, a business can have access to thousands of workers who can, for example, process large sets of data in a relatively short time period. The distribution of workers across global time zones means that the platforms can offer on-demand staffing and task completion on a 24-hour basis. 

By breaking down jobs into “tasks”, platforms facilitate new ways of commodifying labour, and selling it “on demand” to businesses and others who are looking to outsource some aspects of their workload at a lower cost.

Most of the leading technological companies use microtasking platforms for specific aspects of their work, integrating “humans” into what otherwise appears as automatic functions. 

Jeff Bezos famously referred to his platform, Amazon Mechanical Turk as “artificial artificial intelligence” – this is an appropriate description of what is being done and also the importance of human labour for the smooth functioning of technological systems.

Would you agree on it being a new task force that’s also creating a new young tech savvy underclass?

There is nothing inherent to micro-task work that necessitates the creation of an ‘underclass’. Currently the working conditions are precarious, but it doesn’t have to be that way. 

These could be good jobs; it’s just a matter of putting minimum standards in place. At present, the work arrangement is treated as a commercial contract, as the platforms have classified the workers as independent contractors. 

As such the workers are not privy to standard labour protections of an employment relationship, such as minimum pay, leave pay, right to freedom of association, or social security protection, among others. There is also an oversupply of workers on the platforms and intense competition for the available jobs/tasks, which further worsens working conditions. 

What are the working conditions of content moderators, considered the ‘cleaners’ of the internet? Which are the health costs related to the micro tasks they carry out?

There has been some excellent research by Sarah Roberts and others on the working conditions of content moderators. When social media sites were first launched, it became apparent early on that there was a need to moderate content that was uploaded to avoid having pornographic or violent images appear on social media platforms. At the beginning, some of the content moderation was done in-house by the social media platforms, it was later outsourced through employment agencies or outsourced to microtask platforms. 

Now much of the outsourcing is to workers in developing countries (either to a business process outsourcing firm located in the Global South) or through platforms. It is most problematic when it is done through a platform as the worker is at home, isolated, and is likely not aware of what awaits them when they accept the task. There is no support in place for the worker, either by other colleagues or managers, or from in-house psychologists. 

This is hazardous work that has been associated with post-traumatic stress disorder (PTSD) and needs safeguards. 

At zero revenue, content moderation is becoming an increasing financial burden for social media companies who are outsourcing it to countries with cheap labour force and lax regulation. Is this a form of digital colonisation?

The social media companies are doing very well financially, so I wouldn’t say that this is a burden for them – if their websites didn’t filter the images or videos, the success of their sites would be imperilled. As such, content moderation is central to their businesses’ success. I think the best analogy for content moderation being outsourced to developing countries is e-waste: it is mainly generated in the Global North, yet it’s hazardous, no country wants it, and it is being dumped in the Global South. Given the well-documented PTSD effects associated with content moderation, it requires regulation. 

At a minimum, the content should be treated where it is generated (in other words, no outsourcing across borders, and no outsourcing outside the firm).  Regulation would also be effective in forcing social media sites to re-organize the way content is uploaded, so that it’s not so easy for the creator to upload undesirable content. 

Is microtasking the result of a lack of regulation?

Microtasking could – and should — be regulated just like any other economic activity is regulated. People are working on the platforms doing microtask work, just like people work in other jobs. Microtasking came about because the internet needed human intelligence to perform simple functions that computer programmes couldn’t do but that were needed for the smooth operation of burgeoning e-commerce industries, social media platforms and other IT demands, such as training AI systems. 

Microtasking is an efficient way of organizing work, but it doesn’t mean that it can’t (or shouldn’t) be regulated.  If it were regulated – for example, if there were a minimum wage in place – then the platform would have to ensure that the workers’ piece-rate earnings are at or above the prevailing minimum wage. 

This might mean that the platforms would have to allow fewer workers on the platform (so that there is enough work to go around) by screening the workers (similar to when other labour is hired in traditional employment relationships) as well as organize the flow of work to reduce search and waiting times for the workers.  

This would be a positive development. Labour history is replete with examples of how regulation brought about organizational change, usually to the benefit of productivity gains. I believe a similar situation would happen here. At present, there are a lot of inefficiencies in the way microtask work is organized, but as these inefficiencies are only borne by the worker (screening clients to ensure they are reputable and will pay, spending time to understand instructions, taking unpaid qualification tests, etc.), and do not receive the attention they deserve. 

How do the gig economy and microtasking impact the rest of the job market in terms of job conditions and labour rights?

Some of the tasks on microtasking platforms are part of the new digital economy (for example, verifying twitter accounts), other tasks were done previously by on-site clerical workers (for example, audio transcription or data entry) and thus represent a substitution of a clerical, office job (and its associated labour rights and benefits) with a microtask ‘gig’ that, at present, doesn’t entail the same rights or benefits. 

While there are limits to what can be outsourced on microtask platforms, I would expect that in the future more and more work will go to microtask platforms or other platforms including freelancing platforms or “virtual assistant” platforms.  

So much of the discussion around technology and the world of work has focused on technology replacing humans – what we see here and what I believe is the more important trend – is how technology is being used to displace labour to more precarious forms of employment. 

Should governments regulate the micro tasking market and how?

Yes, there is a need for regulation that would impose some minimum standards on earnings, but also ensure that there are neutral dispute resolution systems in place (for example, if a workers feels the need to contest not being paid or being blocked from a platform), and for ensuring contribution to national social security systems. 

However, the cross-border nature of the work – with platforms, clients and workers all potentially located in different countries, makes this task difficult.  

The ILO’s Global Commission on the Future of Work in its 2019 report, Work for a Brighter Future, recognized that the rise of cross-border digital work had resulted in regulatory gaps that required specific interventions at the global level.  It recommended the “development of an international governance system for digital labour platforms that sets and requires platforms (and their clients) to respect certain minimum rights and protections” (ILO, 2019, p.44). It noted that the ILO’s Maritime Labour Convention, 2006 was an important precedent of supra-national regulation, as it establishes and applies a global labour code for seafarers.  Like work on digital platforms, the maritime industry involves multiple parties operating across different jurisdictions.  

An international governance system for digital labour platforms could set minimum rights and protections that would apply to all workers regardless of their location.  It could also impose requirements for the sharing of data between platform operators, regulatory authorities, and the workers themselves that could, for example, facilitate compliance with an established minimum wage. A development such as this would be an important step forward for ensuring workers’ rights in our increasingly integrated world of work.