Capital and Low Paid Digital Data Workers
Over the past two decades digital technologies have made it possible for large numbers of people across the world to find employment with online labour platforms. There are now numerous platforms promoting online work in terms of career opportunities, flexible work options and smart alternatives to non platform employment. Clickworker, Clixsense, The Smart Crowd, Figure Eight and Amazon Mechanical Turk are some of the global players in the rapidly expanding global market.
Amazon Mechanical Turk was established in 2005 and is one of the larger platforms with over 500,000 workers. In the MTurk marketplace clients, misleadingly called ‘Requesters’, post up micro tasks for workers to select and process in exchange for small payments. These micro jobs or micro-tasks have been subdivided from larger jobs so they can be distributed over the internet to a wider pool of labour. For example, the translation of 50 pages by 200 home based workers can be done more cheaply and in less time than that achieved by one worker translating the entire document from beginning to end in an office with overheads. The translated passage might earn the data worker a small payment if it is accepted. The requester pays MTurk a 20% fee on the amount they pay the worker. If they request specific qualifications like fluency in Spanish they pay MTurk additional fees. A per piece wage paid to the data worker (Marx explains piece wages in Capital Volume One) depends on the task being completed accurately within the designated time. While the task may appear to be a simple one, thousands have to be processed swiftly for the worker to earn sufficient to cover the costs of equipment: laptops, computers, smart phones, broadband, maintenance and internet security. Home based workers may also bear the costs of heating or air conditioning, etc. Some platforms charge fees on the a worker’s first payments, and if Paypal is used for transfers more fees will be added to the worker’s production costs.
Microtasking requires human labour because the analytical skills involved in the tasks cannot be done by algorithms yet. Despite all the talk of automation producing widespread unemployment digital technologies have stimulated the demand for online human data workers. MTurk describes micro data tasks as Human Intelligence Tasks. They include information categorisation, data verification, information gathering, copy editing, proofreading, web research, surveys, and so on. The people who ‘perform’ these tasks are not recognised as employees by MTurk. They are ‘Independent Contractors’ rather than employees protected by labour legislation, a minimum wage, pensions, sickness and disability benefits. MTurk declares in its Participation Agreement:
(iii) you will not represent yourself as an employee or agent of a Requester or Amazon Mechanical Turk; (iv) you will not be entitled to any of the benefits that a Requester or Amazon Mechanical Turk may make available to its employees, such as vacation pay, sick leave, and insurance programs, including group health insurance or retirement benefits; and (v) you are not eligible to recover worker’s compensation benefits in the event of injury.Workers perform Tasks for Requesters in their personal capacity as an independent contractor and not as an employee of a Requester or Amazon Mechanical Turk.
Not only do MTurk’s microtaskers have to pay overheads and their own production costs they have no employment rights either. Indeed they occupy a position below employees in MTurk’s employment structure. To process MTurk’s Human Intelligence Tasks they are obliged to accept a subordinate status to employees within the company’s hierarchy.
Many other low paying platforms call their pool of workers ‘participants’ or ‘users’. Some platforms go so far as to redefine the labour process as a contest. ClixSense workers are ‘contestants’ who compete with each other for bonuses and badges like opponents in a game. The Figure Eight platform, formerly ‘Crowdflower’, describes its data workers as ‘data scientists’, irrespective of whether they have studied science and internet technologies. CastingWords keeps control by including a non-disclosure agreement in its package. This obliges workers to agree to not disclosing any information about the company’s products, employment practices and methods of operating. Unlike science platforms such as Zooniverse that bring together volunteers and professionals to identify and classify galaxies, wild animal tracks, wild flowers and much else, digital data labour platforms are profit driven and rely upon keen naive workers willing to comply with strict rules and regulations. Volunteer digital platforms like Zooniverse provide engaging and meaningful work, micro-tasking for MTurk is repetitive, stressful, time controlled and endlessly surveyed. The science crowd worker is part of a community whose members share the same interests. By contrast the home based data worker is divided in space and time from other workers. Their working identity is weakened by being divorced from regulated working practices, regular hours, access to benefits and workplace social rewards.
The redefining of earning workers in terms of data scientists, contributors, participants and contestants is a strategy disguising the reality of the relationship between capital and online data labour. Human labour is still an important source of profits for capital. It is even more profitable if working hours are not hindered by working time directives and complaining Union officials. However the pricing and per piece pay structures of many data labour platforms, the refusal to call microtasking paid employment and the absence of market regulations signify the relationship between labour and capital is far from equal. Some may argue that in order for the relationship to change the power asymmetry has to shift in labour’s direction. Defining ‘labour as data’ might help to restore the balance. This argument misinterprets the logic of capital accumulation, a logic that requires the ownership and control of assets such as data for the production of profits. For example, Lionbridge, the Global communications company behind the The Smart Crowd, tells us it supports integration into eCommerce, Digital Asset management, Marketing Automation, and:
Our 500K+ trainers, raters, and testers in locations around the world make it possible for us to work around the clock, around the sun, to get our customers products to market faster. Are you ready to qualify for work, perform tasks, and get paid? Join our pride
Lionbridge is just one company who rely upon thousands of workers prepared to click around the clock to accelerate the production of profits for client customers and shareholders. Such labour is not free to own and control data that becomes capital. On the contrary workers operate under conditions of discipline and continual surveillance. Internet technologies are changing production forms and processes across the globe but the social relations between capital and labour are more uneven than ever. Fantasies about a cosy re-balancing come unsurprisingly from corporate bosses and their lackeys.
There is nothing new about workers chasing insecure low paying work. They do so because they have no choice, they need to earn enough to pay the costs of living and raising families. In this sense they are dependent upon the market for wages even as they are persuaded to self describe as ‘Independent Contractors’. The US Bureau of Labor Statistics (US spelling) describes these IC workers as ‘contingent’ meaning they work occasionally depending on circumstances; but a job isn’t contingent if it is necessary, that is essential for survival. To avoid destitution workers need to plan ahead, to assess what is necessary.
It is not clear from the BLS’s report how many of the 5.9 million contingent workers in May 2017 were engaged in low paid microtasking and how many of those were women. According to one survey of MTurk workers 69% of the US respondents were female. The author explains this as follows:
‘The main reason for the over representation of females in the US-based workforce is the nature of the tasks and work on Mechanical Turk. Most participants in the US use Mechanical Turk as a supplementary source of income, and often Mechanical Turk is used by stay-at-home parents, unemployed and underemployed workers, and so on.’
The ‘nature of the tasks and work’ on MTurk require just as much focus and concentration as office based work. Women use skills and knowledge microtasking for MTurk clients they could utilise in full time jobs paying the minimum wage. If a higher percentage of them are working from home for low pay while men have regular jobs then that is an expression of gender inequalities in the labour market. The need to supplement wages from low paid caring and servicing jobs probably drives many women to seek income from work that can be dovetailed into their daily domestic routines. Why else do so many platforms promote micro-tasking and online labour with images of youthful women in spacious affluent homes working on expensive computer equipment? Meanwhile 70% of Google’s employed workforce is male, and Amazon, while farming data from online workers, does not release data about its workforce.
It is estimated that by 2027 1 in 3 Americans will depend upon some form of online work to get by, but the US gender wage gap is unlikely to close before 2059. Workers based there will continue to be disadvantaged by gender long after they have transitioned to online platforms. The occupational structures of the US and UK economies are still shaped by differences based on gender, race and ethnic relations even as automation is accelerating the dissolution of non digitised forms of production. Online data workers of the future may find economic inequalities intensifying, making it difficult to access better living conditions and to avoid poverty. Isolated workers under pressure to perform, to keep up, will find it difficult to oppose such intensification. They may not be helped by policy makers who represent the interests of the owners and managers of capital while masquerading as helpers of the unemployed and the underpaid home clicker.
The British Labour Party once agreed
Today the fruit is shrivelled on the plates of home based workers, while the plates of corporate bosses are filled with juicy plumbs, grapes and apples.