Digital work has been hailed as a borderless opportunity for refugees. But new evidence from Kenya, in both camp and urban settings, shows that the digital economy is not an open market. Behind every online job lies a web of human relationships, which include mentors, intermediary organizations and brokers who make access possible.
Across humanitarian and development circles, the idea of digital livelihoods has been promoted as a pathway to refugee self-reliance. Programmes that train refugees for digital or remote work, often linked to digital finance initiatives, are designed to help them earn independently. The underlying idea is simple: with a laptop, connectivity, and the right skills, refugees can “work from anywhere”.
Digital labor platforms such as Upwork and Fiverr have been described as levellers, replacing the barriers of geography with a new meritocracy of skill. Yet for many refugees, this vision remains more rhetorical than real. New evidence based on more than 300 trained refugees in Kakuma and Nairobi (2025) reveals that the key to digital work is rarely the platform itself, but the person—a friend, mentor, an NGO or refugee-led organization (RLO) staff member—that opens the door.
Digital work: Searching hard, finding little
Skilled refugees are not absent from digital labor markets. On the contrary, many are highly active jobseekers who invest time and effort in using digital labor platforms. These platforms dominate their search strategies and embody the hope that digital work can overcome physical and legal barriers to employment. Yet this is where aspiration meets constraint.
In theory, digital work should be frictionless: create an account, advertise skills, and wait for global clients. In practice, refugees face an intricate web of obstacles. Many lack the IDs needed for verification or cannot receive payments due to location, nationality or legal restrictions. Both men and women encounter these barriers, but their impact is uneven. For many women, participation is further limited by care duties, mobility constraints, and insecurity. Travelling to places with stable connectivity is often difficult or unsafe, and limited control over shared devices restricts when and how they can work online.
The findings from our upcoming report on gender differences in skilled refugees’ participation in digital livelihoods in Kenya reveal a persistent conversion gap between visibility and opportunity. Refugees search intensively through online platforms, but few secure paid work there. As one participant explained, “We started with Remotetasks (a microwork platform)… follow up with other platforms… but for now, no job”. Others described how premium subscriptions or ID checks blocked progress.
Not all refugees face these barriers equally. Those with family resources or support from relatives abroad often have better devices and internet access or money for premium accounts, giving them a small but important advantage. For most, however, online markets are spaces of search, not entry. The promise of borderless work therefore masks a hierarchy of access that largely reproduces the inequalities it claims to erase.
Social network analysis helps explain these hierarchies. It moves beyond narrow measures of education, skills, or connectivity to show how relationships and exchanges shape who gains access to digital work. By mapping these networks, this approach reveals the classed and gendered patterns that determine why some refugees can transfer and convert their social and economic capital and secure digital work, while others struggle.
Beyond skills and connectivity, refugees need networks of opportunity
Recognising these barriers is important, since policy debates often frame digital work as a story of individual initiative: gain skills, register on a platform, compete globally. This narrative overlooks the relational labor that makes participation possible. Refugees who succeed are not lone entrepreneurs—they are the outcome of dense social ecosystems.
In our survey, while most respondents said they searched for jobs through online platforms, those who actually secured paid work did so mainly through referrals. These small, often informal acts of connection have become the social infrastructure of the digital economy, both at the local and global level.
Training organizations, spanning RLOs, INGOs and NGOs, private intermediaries, and other actors in the digital livelihoods field, act as key brokers that connect training to income. Many first clients are obtained through these institutions, which vouch for jobseekers’ reliability in markets that often distrust people with refugee status. As one woman explained, her initial contract came about because “they’re the ones who connected me with the job opportunity”. Where verified profiles and client reviews are difficult to build, human credibility becomes the currency of employment.
Yet the ability to draw on such connections is uneven. Our research shows that men benefit from broader and more layered networks that connect them to multiple organizations and peers, offering access to a wider range of opportunities and support. Women’s networks, by contrast, are smaller and less diverse. Our survey data show that more than a quarter of women had no organizational ties, compared with about one in six men. Among those with connections, over half of the women were linked to only one organization, compared with roughly a third of men. These limited and often single-channel connections make women’s networks more fragile and dependent on specific centres or peer groups, leaving them vulnerable when mobility, safety, or organizational support falters.
In Kakuma Refugee Camp in the northwestern region of Kenya, poor and uneven connectivity weakens women’s networks. Many depend on refugee-led centres with timed computer access, limiting chances to communicate and maintain professional ties. One participant explained, “You come at 7:30… at 10:30 you go out, and other groups enter… You work for a few hours, and then you leave.”
In cities, this pattern persists but evolves. Women’s networks remain smaller than men’s, yet within them women take on active brokerage roles. In our data, 72 per cent of female refugees in Nairobi who had done digital work said they had referred others, compared with 56 per cent of their male counterparts. Even within limited circles, women perform the quiet labor of sustaining digital ecosystems that keep opportunities circulating.
Digital participation, for both men and women, is therefore collective rather than individual. Men’s networks are broader and more durable, while women’s are thinner and more easily disrupted. Gender inequality in digital work lies not in the absence of skill but in the uneven architecture of connection that underpins access itself.
Building the social infrastructure of inclusion
Policy thinking often treats digital work as a route to refugee self-reliance, assuming that once infrastructure, connectivity, and digital skills are in place, refugees can independently access global labor markets through online platforms. Our evidence complicates this assumption.
Success depends not only on technical competence but also on social connection: those who can vouch for them, link them to clients, and build trust in markets that often exclude them.
The challenge lies not in refugees’ motivation or ability, but in the structure of digital labor markets themselves. These are not neutral spaces. Entry depends on verification systems, trust networks, and institutional gatekeepers that shape who is seen as legitimate and who remains visible.
To make digital livelihoods genuinely inclusive, investment should focus more on the social infrastructure that underpins access. This includes funding mentorship programs, supporting women-led referral networks, and building income-generating intermediary networks of organizations, such as refugee-led organizations, and social enterprises that remain connected to global markets rather than isolated training pipelines.
Digital economies are often imagined as detached from place and person. The refugee experience makes visible what is true more broadly: behind every digital transaction lies a relationship of trust. It is not platforms that hire refugees, it is people. At a time of shrinking aid budgets and growing calls to localise humanitarian response, refugee-led and refugee-supporting organizations are well-positioned to become central nodes within this network, acting as intermediaries and helping build sustainable, decent refugee livelihoods.
This article was originally published in the LSE Inequalities Blog. Banner image shows digital workers at Action for Refugee Life (AREL), an RLO in Kakuma Refugee camp. Photo: Kimararungu Cadeau Héritier.
