Internet Geographer

Blog

Posts tagged digital divisions of labour
Organising in the digital wild west: Can strategic bottlenecks help prevent a race to the bottom for online workers?


For decades, large firms have been outsourcing and offshoring jobs. Work flowed from from developed economies to developing ones, where wages were lower and regulations were of a lighter touch. Europeans and North Americans lost jobs, and Asians, South Americans, and Africans gained them.

But the nature of these processes meant that business activities had to be on a certain scale to be outsourced. You could outsource a contract for workers to staff a call-desk that expected thousands of incoming calls a day, but you couldn’t easily just get a small website built or a single recording transcribed.

This is now all changing due to the advent of digital work marketplaces. Platforms like Upwork.com, Freelancer.com, and Fiverr.com mediate the auctioning of work. Clients post tasks and workers bid on them, and work essentially becomes a commodity that can be bought and sold in hourly increments.

Some of my colleagues and I have spent the last few years studying this phenomena. We’ve analysed months worth of transaction data from the world’s largest online work marketplace, and interviewed about 150 digital workers in Asia and Africa.

One of the most noteworthy findings in this research is that workers often underbid each other on online platforms: driving down the costs to employers and the wages received by workers. This happens, in part, because there seems to be an imbalance between the supply and demand of work.

Platforms force workers to look at each other’s bids: heightening a sense of competition. Many workers had stories about the invisible ‘other’ - a worker on the other side of the world who could easily take their place if they aren’t competitive.

A Nigerian worker, for instance, told us:

“Most people, most Filipinos, they work for 50 cents. It’s so embarrassing. You’re a professional. You know how to do this job…It’s really affecting those ones that know how to do the job”.

Whilst Filipinos told us similar stories about workers elsewhere. But the net result is the same: a downwards pressure on wages.

What should be done?

In April, I spoke about these issues at a workshop on digitalisation that was put together by the European Trade Union Confederation in Brussels, and the most interesting question that emerged from the resultant discussion was ‘what should be done’? What sort of response could be had by European workers, politicians, and trade unions?

One option is to do nothing. We could accept that if a Rwandan worker does a contract for a British client, then the service is accepted and performed in Rwanda - and thus governed by the labour laws and standards that operate in that country. But to whose benefit is it to allow work to be so easily bought and sold across borders? Employers or workers?

Another option would be to reconsider the very geography of work. We could rethink where the service is actually provided in the case above? Perhaps the case could be made that the service delivered by the Rwandan worker to the British client was carried out in the UK. Perhaps we could at least envision ways to apply minimum UK labour standards (e.g. working time, discrimination law, health and safety, pregnancy and maternity protection etc.) to all contracts issued by UK employers. A UK employer can’t simply treat a UK worker like a disposable commodity, so why should they be able to do so with a Rwandan-based worker?

This sort of strategy becomes more realistic when we realise that although the supply of work in digital labour markets is truly globalised, the demand for it isn’t. Only a few countries in Western Europe and North America are home to the majority of employers. This fact presents us with strategic bottlenecks at which minimum standards can be enforced.

There is no question that this would not be an easy task to undertake, and people who disagree with this vision might point to the fact that much digital work is divided up atomised micro-tasks. Think of jobs that require a worker to write a letter, or fill out a form. It is true that trying to think about what maternity protection means in that sort of context makes little sense. But, it is also true that many digital workers end up working for months or even years for the same employer: and it is for those workers that we need better protections.

Furthermore, we already have laws that govern how to ensure that fair rates are paid. If we can already figure out how workers who are paid on a “per item” basis can earn at least a minimum wage, then surely we can do the same for a digital worker writing blog content.

Other doubters might also point to the absurdity of paying a Rwandan worker a UK minimum wage. Again, this is a valid concern; but does not necessarily mean that we need no standards at all. Couldn’t we, for instance, envision a law that requires employers to pay workers a living wage in whatever country they are resident in?

Finally, the handful of countries home to the most demand for digital work aren’t the only bottlenecks that we can think about as being strategically useful. The platforms themselves are a bottleneck of sorts. All of the big ones are based in the US, Australia, UK or Israel. In the same way that gang masters and temporary agencies are regulated to ensure that their workers are paid minimum wages and given appropriate protections, digital work platforms should also be expected to do the same.

None of this is to say that there is a simple silver bullet to improve working conditions, and it is all too easy to get into a situation in which we pit the interests of workers in some places over those in others. Look, for instance, at the conflict between Western and Central European workers introduced by implementing the EU Posted Workers Directive.

But it is worth remembering that there are now about 3.5 billion people on the internet: most of whom live in low-wage economies, and many of whom are now able to compete for digital jobs. At the moment, the millions of workers who do digital work do so in a largely unregulated way. And while this may benefit those workers who are able to underbid others at the moment, the status quo points to a race to the bottom for the world’s workers: especially as tens of millions of new internet users join the global network every year.

We can never go back to our pre-globalised world of work, but this doesn’t necessarily mean that we should be satisfied with a digital wild west in which atomised workers are left to fend for themselves.  As transnational digitally-exchanged work becomes ever more common, we can change our expectations. And can use what we know about the economic geographies of digital work to envision and hopefully strive towards alternate futures.

Instead of imagining digital work as being undertaken in digital spaces, beyond the realm of regulation, let’s remember it all happens somewhere. Digital work always has a geography.
Kapuścinski Public Lecture - Uneven Geographies of Power and Participation in the Internet Era

Screen Shot 2015-11-10 at 11.50.17

You can watch the whole lecture at the link above. For anyone interested in more about the topic, the following pieces could be of interest:

Graham, M., Straumann, R., Hogan, B. 2016. Digital Divisions of Labour and Informational Magnetism: Mapping Participation in Wikipedia. Annals of the Association of American Geographers. (in press) doi:10.1080/00045608.2015.1072791.(pre-publication version here)
Graham, M. 2015. Information Geographies and Geographies of Information New Geographies 7 159-166.
Graham, M., De Sabbata, S., Zook, M. 2015. Towards a study of information geographies:(im)mutable augmentations and a mapping of the geographies of information Geo: Geography and Environment.2(1) 88-105. doi:10.1002/geo2.8
Graham, M., Hogan, B., Straumann, R. K., and Medhat, A. 2014. Uneven Geographies of User-Generated Information: Patterns of Increasing Informational Poverty. Annals of the Association of American Geographers. 104(4). 746-764. (pre-publication version here)

New publication - Digital Divisions of Labor and Informational Magnetism: Mapping Participation in Wikipedia
Network of edits between world regions, normalised for each target region. The edges are coloured according to the source region. Percentages denote self-edits (not depicted).

I am very happy to announce that a new paper that I have written with Ralph Straumann and Bernie Hogan is now available:

Graham, M., Straumann, R., Hogan, B. 2016. Digital Divisions of Labor and Informational Magnetism: Mapping Participation in Wikipedia. Annals of the Association of American Geographers. 105(6) 1158-1178. doi:10.1080/00045608.2015.1072791.(pre-publication version here)

The paper is the result of about three years of research into geographic patterns of participation in Wikipedia. We sought to ask what the geographies of voice and representation are: with a focus on whether Wikipedia offers people in the world’s economic margins a space to represent their own communities, or whether those margins continue to represented by non-locals. The full abstract is below. I have also previously blogged about some of the results in more detail in the following posts:

Explaining locally-contributed content in Wikipedia about Sub-Saharan Africa

Visualising the locality of participation and voice on Wikipedia

Digging deeper into the localness of participation in Sub-Saharan African Wikipedia content

Informational Magnetism on Wikipedia: geographic networks of edits

Informational Magnetism on Wikipedia: mapping edit focus

Abstract:

There are now more than 3 billion Internet users on our planet. The connections afforded to all of those people, in theory, allow for an unprecedented amount of communication and public participation. The goal of this article is to examine how those potentials match up to actual patterns of participation. By focusing on Wikipedia, the world’s largest and most used repository of user-generated content, we are able to gain important insights into the geographies of voice and participation. This article shows that the relative democratization of the Internet has not brought about a concurrent democratization of voice and participation. Despite the fact that it is widely used around the world, Wikipedia is characterized by highly uneven geographies of participation. The goal of highlighting these inequalities is not to suggest that they are insurmountable. Our regression analysis shows that the availability of broadband is a clear factor in the propensity of people to participate on Wikipedia. The relationship is not a linear one, though. As a country approaches levels of connectivity above about 450,000 broadband Internet connections, the ability of broadband access to positively affect participation keeps increasing. Complicating this issue is the fact that participation from the world’s economic peripheries tends to focus on editing about the world’s cores rather than their own local regions. These results ultimately point to an informational magnetism that is cast by the world’s economic cores, virtuous and vicious cycles that make it difficult to reconfigure networks and hierarchies of knowledge production.

Note that much of this work comes from the following report:

The following paper also offers some related results:

Or for a broader discussion about why the locality of participation matters, see:

"Towards a study of information geographies" A full list of our maps
We very recently published a paper that brings together a lot of the internet mapping work that we’ve been doing:

Graham, M., S. De Sabbata, and M. A. Zook. (2015) “Towards a Study of Information Geographies: (im)mutable Augmentations and a Mapping of the Geographies of Information.” Geo: Geography and Environment, doi:10.1002/geo2.8. (HTML version here)

A more detailed description and abstract are available in my earlier post about the piece. But what I didn’t do there is upload all of the maps and visualisations that we include in the paper. So, here they are. Make sure you check out the (open access) paper if you want to know more about the methods, findings, and implications.

NB. Here’s a talk on the same topic that I did for BBC Radio 4 last year (in case you’d like the non-visual version).