 |
| Valerio De Stefano |
 |
| Janine Berg |
Crowdwork is the casual work of the 21st century. Like other forms of casual work, crowd work is characterized by lack of job security and few, if any, labour protections. But with day labourers, dockworkers, and agricultural hands – probably the types of casual work that most readily come to mind – work is at least for the day. In the on-demand economy, it is for the task at hand. This can be as short as a few kilometers’ drive or ten minutes spent tagging photos on the internet.
Lucas Biewald, the CEO of the micro-task platform Crowdflower, once quipped that ‘before the Internet it would be really difficult to find someone, sit them down for ten minutes and get them to work for you, and then fire them after those ten minutes’. Now instead, ‘you can actually find them, pay them the tiny amount of money, and then get rid of them when you don’t need them anymore’ (quoted in Marvit, 2014).
The ‘Turker’, the Uber driver, or the graphic artist working on an online design platform must continuously search for work, monitoring their computer screens or phones for work opportunities. Indeed, in a recent ILO survey on employment and working conditions in two leading micro-task crowdwork platforms, it was found that workers averaged 18 minutes looking for work for every hour working (Berg, 2016). As an Amazon Mechanical Turk worker from the United States put it,
The toughest part of turking for a living is actually finding the jobs. For every hour I spend working I most likely spend 2 hours monitoring the various scripts I have running to see what jobs show up (Quoted in Berg, 2016:14)
And a Crowdflower worker from Serbia said:
I would very much like if the tasks would open one after another for specific jobs, [rather] than to wait [a] great amount of time for them (Quoted in Berg, 2016:14).
Even when jobs span a few hours or a few days, the worker needs to be constantly searching for new work. Ninety percent of workers in the survey reported that they would like to be doing more work than they are currently doing, citing insufficient work and low pay as the reasons they were not. Despite the desire for more hours, many were already working a lot: 40% of respondents reported that they regularly worked seven days a week and 50% indicated that they had crowd-worked for more than ten hours during at least one day in the past month. Low pay coupled with the need to work resulted in workers spending long hours online.
The survey found that workers averaged between US$2 and US$6 per hour, depending on the micro-task platform and the tasks carried out. Part of the reason for the low hourly rate was the time spent looking for work. Even if a 15-minute task paid relatively well, the time spent looking for it drove average earnings down.
But worker pay is further compromised because this form of work is unregulated. For the most part, the platforms have classified the workers as independent contractors, so they are not privy to the protection accorded to employees on working hours, pay, occupational safety and health, voice and representation, and social protection. This leaves workers taking all the risks on the job. It means there is no floor for wages, allowing earnings to fall below the minimum wage in many of the countries where workers live. Moreover, leave and breaks are not paid, and workers must carry all the costs of social security payments or risk not being covered in the event of disability, job loss or retirement. Indeed, the survey found that only 9.4% of American Amazon Mechanical Turk workers whose main source of income was crowd work made contributions to social security and only 8% made contributions to a private pension fund.
Workers also risk being excluded from fundamental rights such as freedom of association and collective bargaining as well as protection against discrimination, since many jurisdictions reserve these for employees (De Stefano, 2016). Since their right to organize is rarely recognized and sometimes even prohibited by antitrust standards as a form of price-fixing, these workers have an even harder time demanding better working conditions than other casual workers. On top of this workers are dispersed around the world. In most instances, they are alone facing terms and conditions of work unilaterally set by platforms.
In some cases, for example, clients are allowed to refuse payment for an unsatisfactory job while still retaining the work, which may result in opportunistic or illegal behavior. While wage theft is common in other low-wage industries, this rejection feature ‘has effectively legalized wage theft in crowd work, as there is no way to distinguish between wage theft and legitimate and normal use of an intentionally designed platform feature’, (Silberman and Irani, 2016:518).
Workers can also be excluded from the platforms and apps or be prevented from accessing better-paying jobs on the basis of negative ratings. Review systems expose workers to implicit or explicit discrimination (Leong, 2014). Furthermore, ratings and reviews are one-sided: workers are seldom allowed to review clients or respond to the feedback. In response, some workers have organised their own forum and methods for providing ratings of clients, as in the case of Turkopticon, a plug-in for Amazon Mechanical Turk, that “helps the people in the 'crowd' of crowdsourcing watch out for each other – because nobody else seems to be” (Turkopticon, n.d.).
Besides reviewing and evaluating workers’ performance, platforms are also very effective at monitoring what workers do. Upwork, the online freelance marketplace, offers its clients the option of paying by the hour, as it can monitor workers by recording their keyboard strokes and mouse clicks and taking random screen shots. Uber expects drivers to have the app on and prolonged periods without logging on can lead to account deactivation. The app tracks drivers’ whereabouts even during their downtime. Drivers are expected to accept the rides the app assigns to them. If they cancel or fail to accept as few as 10% of rides, their account may be deactivated, and the worker essentially dismissed.
But the same technologies used to monitor workers could be used to protect their rights. The platforms know how much time workers spend online searching for jobs, they know when they are working and taking breaks, and the quality of their work. Why can’t this same technology be used to monitor working time to pay a wage that at least complies with the minimum wage and enable social security payments? Why can’t they use the technology to better organize the work so that workers’ search time is minimized?
The platforms will not self-regulate to offer better working conditions. And well-intended platforms will have trouble surviving in what is a global race to the bottom. Unless governments step in and recognize workers as the employees that they are, platforms will continue to have an advantage over traditional industries, risking a deterioration of working conditions that extends beyond on-line work. With nearly unlimited supplies of labour and an absence of liability placed on platforms, casualization will continue. As one respondent in the survey mentioned above noted, “This is obviously a way of working that will likely explode in the future. If some sort of fairness were present in early stages it would prove beneficial to long term prospects.”
Janine Berg is an economist with the International Labour Office.
Valerio De Stefano is a lawyer with the International Labour Office.
References
Berg, J. (2016) Income security in the on-demand economy: findings and policy lessons from a survey of crowdworkers, ILO Conditions of Work and Employment Series, Working Paper No. 74, Geneva, ILO.
De Stefano, V. (2016) The rise of the just-in-time workforce: on-demand work, crowdwork and labour protection in the “gig-economy”, ILO Conditions of Work and Employment Series, Working Paper No. 71, Geneva, ILO.
Leong, N. (2014) The sharing economy has a race Problem, Salon, 2 November.
Marvit, M.Z. (2014) How Crowdworkers Became the Ghosts in the Digital Machine, The Nation, 5 February.
Silberman, S., and Irani, L. (2016) Operating an employer reputation system: lessons from Turkopticon, 2008–2015, Comparative Labor Law & Policy Journal, 37 (3) Spring.
Turkoptikon (n.d.)
NB: The views expressed in this column are the authors’ own independent views
0 Comments:
Post a Comment