As part of our project on digital labour and development, Isis Hjorth, Mark Graham, Helena Barnard and I have been meeting and interviewing over a hundred people who do freelance work over the Internet, through platforms such as Upwork and Elance. One of the big impacts these platforms are expected to have on the economy is disintermediation: that buyers and sellers can do business directly, without companies and agencies in the middle. Gatekeepers disappear and hierarchies melt down into market-based one-to-one connections that can span the globe. This idea is sometimes referred to as the “peer-to-peer economy”.
Our preliminary results suggest that disintermediation is indeed happening. People from Manila to Cape Town and Saigon to Nairobi work for over the Internet for clients all over the world. But we are finding that the story doesn’t end there: in many cases, new network patterns emerge to complicate the picture. Some contractors take work from a platform and then pass it right back into the platform at a lower rate, reintermediating it. Some pass online work to workers they have hired from the local labour market, lengthening the value chain locally. Others collaborate horizontally with family and friends to complete their projects. None of this is visible online: all you see on the platform are faces of individual contractors. But behind a face there is often a small network of people organized according to one of a variety of patterns.
“[E]verybody is trying to get a piece of the commission but they don’t want to do the hard work […] It has happened to me that my employer and I bidded on the same project. The employer gets awarded the project and it gets passed to me for a lower fee.” – female, 28, Malaysia
In the reintermediation pattern, a contractor who wins a job on a platform turns around and re-lists that job on the same platform with a lower reward. The actual work is thus done by a sub-contractor, but the contractor captures a slice of the revenue. Some contractors are able to do this because they have a very attractive profile on the platform, helping them to win projects. A long track record of successful projects and good reviews are very helpful in convincing clients. This “platform capital” allows successful contractors to insert themselves in the chain between clients and less qualified contractors.
Reintermediators are not necessarily just rent-seekers who capture value without creating any themselves. Reintermediators can provide value to the client by using their superior information of the platform to find good subcontractors. Reintermediators can also reduce the client’s risk by ultimately doing the work themselves if the subcontractor fails. This is because the reintermediator can’t risk a bad review, positive platform capital being their main asset.
“I wear the proofreader’s hat, editor’s hat, whatever else that’s left and they [local subcontractors] do just the raw transcription. […] I look at myself as someone who is employing other people, training them to do ethical things.” – female, ~30, Philippines
In the local lengthening pattern, a contractor takes work from an online platform and passes it on to workers they have hired from the local labour market. These can be either subcontractors or actual employees. The contractor is able to do this because of their superior knowledge of both the online platform and the local labour market. The foreign clients in the platform have no way of directly accessing the local labour market in, say, Philippines. And people in that local labour market don’t necessarily have any way of accessing online platforms, for example due to language barriers or simply lack of awareness. The woman who provided the quote above used the local equivalent of classified ads to hire freelancers for audio transcription projects she obtained from clients online.
Besides simply benefiting from their information and access advantage, the contractors who perform local lengthening can be real managers who add lots of value to the process. They can take a large project from the online client, split it into smaller parts, find a suitable worker for each part, supervise each part, integrate the results into a final deliverable, and ensure timely delivery. Though platforms reduce information and coordination costs, they don’t eliminate these costs completely, meaning that small hierarchies can still often be more efficient than purely one-to-one market relationships.
“Sometimes I can’t work, like I’m busy maybe I’ve gone upcountry and stuff like that, I leave him my account, he’s able to manage it and communicate with clients and go on. […] We have become like brothers because of this […]” – female, 37, Kenya
Under the horizontal collaboration pattern, we placed a variety of non-market relationships between online workers and their family and friends. In some cases, family members worked on online projects together and shared the income. For example, the face appearing in one online profile belonged to a lady with attractive features and good language skills, while much of the production work was actually performed by her little sister. In other cases, friends who both had a profile in the platform (or in different platforms) sent each other work without money changing hands. The key point here is the use of social capital developed outside work to establish and maintain mutually beneficial economic relationships; so-called multiplex ties.
These are just preliminary findings. Our empirical research and especially the analysis of the findings is still very much ongoing. But some things are already clear. While digital platforms can disintermediate economic relationships and deconstruct old hierarchies, that’s not the end of the story. New structures of hierarchy and collaboration can emerge both at the edges of the platforms (as in local lengthening and horizontal collaboration) as well as inside the platforms themselves (as in reintermediation). These structures can be responses to the information and coordination costs that still remain, but some are also responses to the new incentive structures that the platforms themselves inadvertently create, like the value of 5-star ratings. We will continue to study what these patterns are, why they arise, and what impacts they have on the nature of work and everyday life.
We are presenting these preliminary findings at the American Sociological Association’s Annual Meeting in Chicago this Sunday at 8:30am. You can download the paper here:
Lehdonvirta, V., Hjorth, I., Graham, M., & Barnard, H. (2015) Online Labour Markets and the Persistence of Personal Networks: Evidence From Workers in Southeast Asia. Presented at the ASA Annual Meeting 2015 session on The Changing Nature of Work, Chicago, 22-25 August.
Note: This post was originally published on the OII's Connectivity, Inclusion, Inequality blog on . It might have been updated since then in its original location. The post gives the views of the author(s), and not necessarily the position of the Oxford Internet Institute.