Our Programmes



Sign up to our newsletter.

Follow ODI

What Kenyan ghost-writers can teach us about prejudice in the digital gig economy

Written by Susan Njambi-Szlapka

Would you trust a Kenyan to write your Bachelor’s or Master’s thesis? Recent research finds that Kenyan nationals are not trusted by American students to do so, forcing academic ghost-writers to hide their identity to win work.

Besides obvious ethical issues, this raises wider questions about the online gig economy including exploitation, trust and discrimination, including how we talk about equality in digital labour.

Coronavirus has also fuelled a sharp decline in demand for and an increase in supply of online workers. Online “place-based” platforms that require physical on-site delivery such as Deliveroo or GetBoda are seeing an increase in demand. But online gig workers on platforms like UpWork are facing more competition for fewer jobs as clients curb non-essential spending and more people seek digital sources of income. These same platforms are even asking online workers to pay in order to bid for work.

Under these circumstances it is crucial to understand how online gig workers in developing countries, who often have no access to any form of social protection, compete in the online gig economy to earn their livelihoods while identifying ways to combat the discrimination many experience.

One account, many identities: the online gig economy labour market

There are 6 million people in developing countries participating in online labour according to the Online Labour Index, while 60 to 70% of all online gig workers in Kenya are ghost-writers. However, this wrongly assumes accounts on gig economy platforms like UpWork have just one user per account.

Instead, accounts with good reviews are often shared between ghost-writers and traded on the black market, making ghost-writers key contributors to this underreporting. Owners of such accounts employ others to do the actual work, while using VPN’s (Virtual Private Networks) to hide user’s locations, making many ghost-writers in Kenya and elsewhere invisible.

Therefore, the real number of online workers is estimated to be up to 8 times higher. So why is this behaviour so widespread?

The false promise of digital globalisation

In the flat-world narrative of globalisation, online work is based on merit and value for money, as technology levels the playing field between workers in the Global North and South by removing the impediment of geography. This apparently allows workers in the Global South, as long as they have internet access and possess the right skills, to outcompete their northern counterparts by underbidding for online work generated in a small number of rich countries in the Global North.

However, recent studies find that prejudice around geography and identity are still key drivers in who wins work and under what conditions. Young ghost-writers in Nairobi even report that advertised jobs sometimes go as far as saying that people from certain countries should not even “bother bidding”.

Reputation is everything, but it is more costly for Global South workers

Cognitive shortcuts have been well-established in hiring practices on the platform Nubelo. Clients assume workers from lower income countries will provide less valuable work and therefore pay them disproportionately little or don’t hire them at all. In the absence of organisational affiliation, workers from developing countries are perceived as representatives of a nationality, rather than individuals with different experiences, credentials and personalities.

The exception is where a worker has a good reputation.

The reputation of an online worker, captured through ratings by previous clients, is key for their competitiveness and is much more costly for those in lower-income countries when it impacts their likelihood of employment.

However, if a client has a good experience with someone from a developing country, they are more likely to hire other workers from that same country. Therefore, while in the short-term hiding identity and location helps, it could inhibit worker’s future long-term ability to find work.

Four ways we can mitigate online discrimination

  1. Online platforms can learn from analogue labour market measures to enable merit-based hiring practices (including for instance anonymised CVs).
  2. More rigorous research could help us understand what is exacerbating discriminatory hiring practices including language barriers, sense of shared values and expectations around education systems.
  3. Increasing fairness through algorithms could include prioritising displaying information relevant to a person’s qualifications or credentials, rather than their location (this could even go further and place people from certain regions at the top of search results). Nubelo has already used this to help newcomers compete against established online workers.
  4. Finally, increasing awareness among clients hiring online workers could also help address these biases.

How we navigate the world of online work

There is still a global division of labour determining what type of jobs online workers in the Global South are entrusted with and those they are presumed unqualified for. This division, based on prejudice, will likely reinforce old power structures where repetitive or less complex tasks are left to the Global South, while complex and better paying tasks remain in the Global North. However, as the former become automated, this could leave workers in low-income countries vulnerable.

If demand for online workers is diminishing in the wake of Covid-19, the situation in developing countries will only worsen. Governments, donors and online platforms in the north and south will need to devise ways to overcome these issues of discrimination, while also supporting online gig workers both now and post-Covid.