Author: Niels Beerepoot

Digital Economy, Digital Labour

Digital labour platforms as winner-takes-all markets: implications for fair work standards

Niels Beerepoot

A number of contributions in the DIODE blog series have focused on how digital labour platforms such as Upwork and Freelancer provide new economic opportunities for online freelancers, particularly in the Global South. Subsequently, relevant questions have been raised on the precarious nature of digital gig work (see e.g. blogs by Balaji Parthasarathy, August 2018; Richard Heeks, July 2018; Mark Graham, December 2017), and how the regulatory framework for platforms can be strengthened to guarantee ‘fair work’ principles for those active on digital platforms. Initiatives towards third-party monitoring and certification in this field, and how decent work standards can be met, are interesting and should receive broad-based support. However, through the very nature of their functioning, digital labour platforms will always be plagued by a level of unfairness, which will be discussed in this blog-post.

One central but largely unexplored topic in the literature on digital labour is what it takes to be successful on digital labour platforms. As its central features, the competition for work is global, employment relations are flexible, and individuals rely on reviews and performance scores for securing new jobs. In practice, financial gains on digital labour platforms are highly uneven with a small number of freelancers securing the majority of the jobs. In management circles this would be regarded as the 80/20 rule (pareto principle), which in this case refers to the situation that 80 percent of the work is likely done by only 20 percent of the freelancers that are registered on the platform. In a similar vein, digital labour platforms are a typical example of a winner-takes-all market. The concept winner-takes-all markets was originally developed by Frank and Cook (1995, 2013) to highlight the emergence of top-earners in particular professions such as sports, music or other creative professions. In the digital economy the winner-takes-all principle can be identified at two levels: the digital platform industry at large, where platforms such as Facebook, Uber, Amazon.com and Upwork currently dominate their market niches, and for those working via platforms. Formal entry barriers to digital labour platforms are low, which means that anyone literate with a computer and access to the Internet can sign up. However, successful participation in digital labour platforms is a complex game because of the diversity of actors involved.

Whereas in local or national labour markets most buyers and sellers of labour share common grounds in terms of cultural background, education, and their understanding of what represent good work skills, work ethics and decent pay, this is not true on global digital labour platforms. Here, clients posting a job are likely to be presented with freelancers that are based, raised and educated somewhere half a world away in a profoundly different socio-cultural and educational context than the client’s. This creates an assessment challenge for clients, and adds a measure of uncertainty to hiring transactions. An additional measure of uncertainty is created by the digital nature of the working relation between clients and contractors. With clients and contractors at a distance and chances of them ever meeting in person being slim, the embedded relations and social monitoring and sanctioning mechanisms that encourage people to perform well and live up to their promises in offline working environments carry less weight, making it easier for either freelancers or clients to disengage or otherwise fail their counterparts. Under such conditions, contractors with good credentials (or a good digital reputation) are more likely to make it to the shortlists of clients looking for someone to hire. They are also more likely to be rehired and be able to secure higher wages for their ‘proven’ trustworthiness.

Skills-tests via digital labour platforms are the most obvious way for new entrants to show their capacity and to reduce uncertainty of clients. Surprisingly, Upwork (the largest global platform for digital work) announced in July 2019 that it will remove the various skills tests on offer via the platform, as clients didn’t find skills tests important when making hiring decisions. Rather, according to Upwork, clients consider profile introductions, portfolios, and job feedback a better showcase of a freelancer’s skills and experience. This shows the very nature through which digital labour platforms evolve for freelancers into winner-takes-all markets as it favours the more established freelancers with good credentials (for example a high performance rating and positive feedback). When client-generated feedback constitutes an important part of contractor credentials and is used by clients to (largely) base their hiring decisions upon, it favours contractors who have already received good feedback and it puts newcomers that have not yet had their chance to accumulate good feedback and build a digital reputation at a disadvantage. Such effects are likely to be more pronounced if clients in their assessment of potential contractors, have a strong interest in qualities that can be demonstrated only ‘on the job’ and not easily upfront. This increases clients’ reliance on the feedback received by contractors from other clients, and thus reinforces the importance of the feedback loop. This makes the platforms prone to exclusionary effects (in addition to distribution effects) as newcomers will have a hard time establishing a foothold in the market.

To conclude, the greatest contradiction of digital labour platforms is that success is very visible (e.g. through information on profile pages of freelancers) and newcomers are often lured to join (see blog by Niels Beerepoot, September 2017) but securing a position in this market is a major challenge. New entrants need to overcome barriers of uncertainty and lack of a digital reputation. These particularities in the functioning of digital labour platforms are hard to integrate when defining fair work standards but to a large extent determine how freelancers perceive digital gig work. Fair work standards are now oriented towards those who have acquired a seat at the table where joining the table is the hardest task.

Frank, R. & P. Cook (1995). The Winner-Take-All Society: Why the Few at the Top Get So Much More Than the Rest of Us. New York: Martin Kessler Books
Frank, R. & P. Cook (2013) Winner-Take-All Markets. Studies in Microeconomics Vol. 1 (2): 131-154. Continue reading “Digital labour platforms as winner-takes-all markets: implications for fair work standards” →

Digital Labour

The changing strategies by digital labour platforms and its impact on freelancers in the Global South

Niels Beerepoot

Policy makers in the Global South have discovered digital labour platforms (e.g. Upwork, Freelancer) as a potential source for employment. In his blog-post of June 28, Brian Nicholson mentioned initiatives in Pakistan and Nigeria that train young people for online work. In the Philippines, the Department of Information and Communications Technology (DICT) organises so-called ‘rural impact sourcing’ workshops around the country. Despite using the term impact sourcing, these events have no connection with the original conceptualisation of impact sourcing which highlights how social enterprises hire and train marginalized individuals for digital work. These events are often sponsored by the country’s main internet providers and aim to make participants interested in becoming ‘Online Filipino Workers’ (as paraphrase of the well-known term Overseas Filipino Worker). Apparently, for policy makers in the Global South, the digital labour market is seen as an infinite source of employment that should be tapped rather than restricted or regulated. Such workshops also still present digital labour platforms as ‘global labour auctions’ in which everyone can participate regardless of his/her location as long as there is digital connectivity (for which an internet provider that you can sign up with is right at hand).

In reality, digital labour platforms (like in this case of Upwork) no longer function as a labour auction. In the past few years a number of measures have been taken by Upwork through which it increasingly benefits and favours a small number of successful, high-earning freelancers. The ease of entry for freelancers had created a situation in which even for jobs with limited remuneration dozens or even hundreds of freelancers would apply. With the platform still retaining a fixed percentage of the agreed remuneration as its fee, the so caused downward pressure on price formation also ate into Upwork’s revenues. Obviously, from a business perspective, it is not in Upwork’s interest to have too many freelancers competing on the platform, causing a downward pressure on price formation and the gains from the work performed to be spread too thinly. The introduction of a $3 per hour minimum wage in 2014 had been a first step attempt at ‘price support’ and additional value capture by Upwork. However, in practice it led to more projects being offered for a (low) fixed amount instead of for an hourly compensation. Therefore, Upwork needed a more drastic shake-up of its business model in order to generate more revenues for its investors.

Upwork’s transformation of recent years is best summarized as a change from labour facilitator to labour arbitrator (see Govil and Patnaik, 2014). The platform now no longer only provides a meeting place for the supply and demand of online labour but is directly involved in the recruitment of freelancers for clients and also handles pay-rolling (through Upwork Payroll). In 2016, it introduced Upwork Pro and Upwork Enterprise. In Upwork Pro clients receive help with drafting the project requirements and Upwork selects and shortlists premium freelancers for the job. Upwork Enterprise is aimed at larger organizations and involves the full-scale management of freelancers by Upwork. Upwork’s commission is no longer a flat 10 per cent, but varies with the size of the project to incentivize higher-value, longer-term projects. The main change for freelancers is that they can ‘buy’ their visibility in the search machine by taking a paid membership of $10 per month, which obviously benefits the financially stronger freelancers. An additional measure is the ‘job success rate’ of freelancers, an algorithmic score that apart from the ‘old’ job performance rating also includes such factors as whether the freelancer has won repeat and larger contracts, and experienced disputes with clients. It clearly favours freelancers who have managed to land longer term collaborations with single clients over those who tend to earn from multiple small gigs (which is more common among freelancers working part time).

Successful freelancers are now invited to join a talent pool which connects them to larger client projects. A catch is that those who accept, are required to be available for Upwork only (committing at least 30 hours per week). On the other side (and seemingly inspired by Taylorist management principles), Upwork has started to suspend accounts of freelancers who lack distinctive skills or who are not generating enough revenues. New freelancers (particularly those with ubiquitous skills) are also no longer automatically accepted on the platform. This means that Upwork now more actively intervenes in the pool of labour it offers to clients, making freelancers vulnerable to sudden policy changes by the platform. Once freelancers work via the platform, Upwork expects their loyalty and exclusive, full-time availability. However, Upwork does not offer any kind of social protection in return. Upwork tries to control (and discipline) its labour pool through negative measures (such as the fear of terminating the profile or reduce the visibility in the search machine). Here you can wonder how well aware they are of the everyday lives and responsibilities of their freelancers in the Global South.

This blog showed the contradiction between Southern policy makers who (still) view digital labour platforms as an infinite source for employment, and the platform which is tightening the grip on its pool of freelancers and which makes it increasingly hard for new entrants to join. Recent measures by the platform work against freelancers from the Global South, particularly the many of them working in the low-skill segment of the digital labour platform. By providing more services and introducing more rules, platforms like Upwork increase their influence on how freelance labour is globally traded and, in the process, try capturing a larger share of the value that they help to create. Upwork’s evolution from a discrete facilitator towards an ever more manifest, some may say obtrusive arbitrator, shows that in fact a new intermediary, equipped with new measures for disciplining labour, has been born. Not only does this require new legislation, it also requires a debate with platform owners with regard to their responsibilities towards their community of workers.

Govil, A. and S. Patnaik (2014). The Future of BPO: How Human Cloud and Infrastructure Cloud are changing the Game. Los Angeles: Avasant Research http://avasant.com/insights/digital/the-future-of-bpo-how-human-cloud-and-infrastructure-cloud-are-changing-the-game/