Digital infrastructure projects are mainly developed in the Global North and it is therefore unsurprising that these projects address issues relevant in developed countries. It is critical, however, to understand the barriers to uptake and production of digital infrastructures in Africa, specifically open-source development of infrastructure.
In 2021, Research ICT Africa was awarded a digital infrastructure grant to evaluate the landscape of African digital infrastructures; and investigate how African participation in the development, maintenance and application of global open-source digital infrastructure can be enhanced. The project investigated the following topics:
- The extent of African contributions to the codebase of critical digital infrastructures.
- The challenges, constraints, and obstacles to the development of critical digital infrastructures on the African continent.
- The lessons that can be learnt from successful African-led open-source projects.
- How, if at all, digital infrastructures are different in the African context.
The research study consisted of four components: a literature review, a quantitative study of African participation in-open source digital infrastructure development, a qualitative study focusing on African-led open-source projects, and an open-source policy case study.
The qualitative analysis of the extent of African contributions to the codebase of critical digital infrastructures included case studies of well-known flagship projects and initiatives on the continent as well as some new lesser-known projects. Some of these are iNethi, Village Telco, Grassroots Economics, Ushahidi, Masakhane, OpenUp, Crane Cloud, Open Source Community in Africa, and Translate.org.
Open-source code is partially produced by corporations for economic incentives, but also by individuals who are able to contribute for other incentives including problem-solving, mastery of technology, status within an expert community, and a desire to benefit others. However the additional capacity to engage in social production is dependent on available time, energy, and access to a networked computer (Benkler, 2007). In Africa there are simply fewer technology corporations and fewer individuals with the available capacity to contribute to open source.
So, what are the challenges, constraints, and obstacles to the development of critical digital infrastructures on the African continent?
First of all, there are funding challenges – for many of the projects we analysed, the inception was challenging because it was hard to get funding without proof of prior experience or a history of grants previously received. On many occasions, organisations and individuals relied on personal funds or support from family and friends to prove that their ideas were worth investing in.
Our funding has come mainly from donations from family members and humanitarian organisations that have seen the effectiveness of our work. Note that our budgets for technology are never funded for the sake of tech, but rather for specific humanitarian applications.Will Ruddick, Grassroots Economics, Kenya
In the African context – similar to Information and Communication Technology for Development (ICT4D) projects – the financial sustainability of open-source projects largely depends on sustainable business models. Many open-source projects are donor funded, and there is often a lack of vision on sustaining an initiative once the funding runs out. On some occasions, there was a mismatch between donor interests and an organisation’s motivation to develop open-source solutions, which caused strains in communication. For example, some organisations or governments may fund an initiative but are not willing to make the data openly available.
A commonly adopted model by African non-profit organisations deploying open-source infrastructure is donor funding, complemented by providing consultancy services on either tools they build or subject matter expertise.
The open-source movement generally has a complex mixed set of motivations for participation. Some funded initiatives are only able to attract the developers that they need by paying a market rate for contributions. Other initiatives rely on voluntary work from contributors whose main motivation is often to build their portfolio and reputation. In the case of African projects, code is largely written by students as part of their studies or as they prepare to get into the job market, which makes regular contributions challenging. On the other hand, many contributors from the Global North decide to participate as volunteers, driven by philanthropic motivation as well as the desire to make an impact.
In Africa, the majority of well-established open-source projects in the application space are from South Africa and Kenya. According to an interviewee, the average age of Kenyan developers is 23 years, which is extremely young. However, unlike many software developers in the Global North, coders usually start doing personal interest projects and open source at a later stage of their career, often once they have accumulated some money and confidence to pursue their own interests. This might explain why the potential pool of African contributors is relatively small.
In Kenya, once you’re in IT, you are the hope of your entire family and need to send money back home. The tech space is growing, so young developers are very likely to land a corporate job and make a living, rather than hacking around with open source.Freelance developer, Kenya
There were other interesting insights on data.
While contributions are an important aspect of maintaining open-source projects, some organisations are continuously working to ensure that data is not only open but is of good quality and usable.
Additionally, one of the research participants spoke of knowledge gaps as a barrier, such as not having a full understanding of open source and getting fewer contributors from non-English speaking countries on the continent.
Diversity is still a major problem in the continent, particularly for example within our organisation we haven’t seen anyone from the francophone region. Language barrier is something we are hoping to solve. Even within Northern Africa, Arabic is something we need to work on […] we still have a long way to go.Samson Goddy, Open Source Community Africa, Nigeria
While all these challenges persist, there still are some quite successful African-led open-source projects. The quantitative study of this project reveals interesting insights into what makes a project “African” and who is contributing to the projects. The next couple of blogs from our project will go into detail about contributions on GitHub and African authors or co-authors who have contributed to internet standard development.