In Search of Flexibility

The freedom and flexibility to work at any time and from any place is often heralded as one of the benefits of microtask platforms such as Amazon Mechanical Turk, Clickworker, etc., along with easy accessibility of work itself (Barnes, et al., 2015; Felstiner, 2011). This flexibility allows workers with care and household responsibilities and with health problems to perform tasks from home.

Micro work sounds appealing in practice. However, access to these online platforms can be quite complex. Workers invest a lot of time in building their profiles, performing tests and unpaid tasks in order to earn qualifications, and only then can they access paid tasks on these platforms.

In addition, platforms can have explicit entry barriers, wherein workers from certain countries simply cannot register. The turnover rate on microtask platforms is also quite high, as many workers discover over time that work comes their way only if they are skilled and diligent. Moreover, much of the available work does not fulfill worker expectations with regard to interests or earnings.

The flexibility on digital labor platforms is, thus, quite illusory. Workers cannot really log in from any location, since the tasks they want to perform may not be available at that precise time and place. It has been estimated that on any one platform there are about 100,000 to 600,00 tasks available at any given time. However, since the global pool of workers on these platforms is significantly larger, the competition for these tasks is high. As a result, the availability of tasks is insufficient and inconsistent.

For instance, in our global survey, a large proportion of workers (88 per cent) reported that they wanted to do more tasks related to micro work than were available. About 58 per cent of the workers reported that the availability of tasks was insufficient, and an additional 17 per cent indicated that they did not find enough well-paid tasks.

The reason behind the insufficiency of work can also be linked to the platform design and result in discrimination. Some platforms provide features that allow clients to select who can access the task- between the global pool of labor or by specific populations, based on certain characteristics, such as place of residence, gender, age, etc. In such instances, it is possible that some of the well-paid tasks do not reach workers in certain countries or certain groups, such as women, older or younger workers, etc.

Additionally, workers spend many hours of unpaid work on digital labour platforms. Workers reported working 24.5 hours doing micro work in a typical week. Of this, 18.6 hours was paid work and 6.2 hours – around 25% of their micro work time, was unpaid work. That is, looking for tasks, performing qualification tasks, completing tasks for which they were not paid, etc.

On average, women perform fewer hours of paid work, though the number of hours spent on unpaid tasks is quite similar to that of men. The irregularity and insufficiency of work, has led many workers to work long hours.

The intensity of work is quite high as well, as 44 per cent of workers reported that they worked for more than 10 hours per day for about one-third of the month (1-10 days), and 23 per cent of the workers worked for more than 10 hours per day for two-thirds of the month (11-30 days).

The need to constantly look for work, as it is not regularly available, leads many workers to work unsocial hours. Of the workers interviewed, about 43 per cent reported working during the night (10pm to 5am) and 68 per cent reported working during the evenings (6pm to 10pm), due to the temporal distribution of jobs.

Irregular Work: Low Earnings and Precarious Conditions

Apart from flexibility, a number of platforms entice workers by advertising how much they can earn per hour. The ILO survey reports that on average, in 2017, a worker earned USD 4.43 per hour when only paid work is taken into consideration. This figure reduces further to USD 3.29 per hour when both paid and unpaid work is accounted for, ranging between USD 1.92 and 6.54 across the different platforms.

Despite their high education levels, a substantial proportion of workers earn below the average wage per hour (70 per cent of workers in CrowdFlower, Clickworker, Microworkers and Indian workers on AMT) and their local minimum wage (over 60 per cent of American workers earned below the federal minimum wage of USD 7.25), Low remuneration is due to rejections (whether justified or not), errors in the task, the way the platform is designed, and wage theft.

For instance, about nine out of ten workers in the ILO survey reported that work was rejected or that they were refused payment- which amounts to wage theft. The workers voiced frustration as they were unable to appeal for unfair rejections.

About nine out of ten workers in the ILO survey reported that work was rejected or that they were refused payment- which amounts to wage theft.

In addition, there are differences in pay between workers from different countries even within the same platform. For example, on Amazon Mechanical Turk (AMT) the average hourly earnings of American workers are about 2.5 times that of Indian workers. About 71 per cent of the Indian workers earned less than USD3.50 (the Indian average), while only 5 per cent of the American workers earned less than the Indian average. The median wages of a “typical” Indian worker on AMT was USD2.14, compared to USD7.50 of an American worker.

A part of the earning differential is due to the nature of tasks – as the best paid tasks, like content creation, are often only available for the American workers. Low-end and low-paying tasks are left to Indian workers.

The global outsourcing of work through digital labor platforms has led to the development of a 24-hour economy.

These empirical findings from ILO’s global survey reveal that the freedom and flexibility of microwork is quite illusory. The global outsourcing of work through digital labor platforms has led to the development of a 24-hour economy.

ILO’s findings show that while these platforms provide a number of opportunities to workers, they also expose them to risks, relating to their basic rights and working conditions.

This apart, many of the platforms charge workers a fee for performing tasks on the platform. This is in violation of the ILO’s Private Employment Agencies Convention, 1997 (No. 181), which clearly states that, “private employment agencies shall not charge directly or indirectly, in whole or in part, any fees or costs to workers”.

Finally, there is no mechanism for workers to raise their concerns relating to payment for tasks, undue rejection, harassment or other unprofessional conduct. Nor is there any formal mechanisms available for workers to engage in collective bargaining. As these platforms operate in a regulatory void, they violate some of the core ILO Labour Standards and Conventions.

In the absence of crucial labor protection policies, working conditions for micro workers are likely to remain precarious under the garb of ‘flexibility’.

The research findings of this article are based on a recently published ILO report titled Digital labour platforms and the future of work: Towards decent work in the online world, which provides a comparative study of working conditions on five microtask platforms based on a survey of 3500 workers.

Micro work refers to small clerical tasks which can be performed remotely. It includes tasks such as image identification, transcription and annotation, data collection and processing, content moderation, etc.