Executive Summary

This study builds on the notion that IT-enabled services exports – the types of digital work that requires only a computer and an internet connection and can thus be conducted from anywhere in the world – can become a promising sector for future-oriented job creation in Africa, and a driver of catch-up development. While IT-enabled services exports are still in their infancy, they seem to be a legitimate contender for job growth amidst the digital era when compared to more traditional routes to prosperity – agriculture, manufacturing, or domestic services (Melia 2020). If this is true, then questions of how and where growth of this sector might unfold become interesting for many different actors: from global offshoring companies to local firms, African governments and training institutions, bilateral and multilateral partner organisations – all have a stake in the growth of a healthy IT-enable services sector. The current study assesses the sector by examining the types of tasks that are already being exported from Africa, and different types of locations in which clusters of new IT-enabled service providers could emerge and grow. I take on the lens of aid-driven organisations that seek to boost the sector – since one such organisation, Digital Skills Accelerator Africa (DSAA) commissioned this research. But the analysis and policy recommendations are of a general nature and can hopefully be useful for a wide audience.

The study is organised into 6 sections that successively grow in length and importance: two short sections on the Literature and History (Sections 2 & 3) are followed by two longer sections on the Tasks and Locations of IT-enabled services in Africa (4 & 5). Finally, the largest section (6) pulls together the strings for Policy Recommendations for flourishing IT-enabled services export sectors in Africa.

 Literature and historical context of IT-enabled service clusters

Clusters of economic agglomeration have driven new tasks and sectors since the First Industrial Revolution. The science of where such clusters emerge shows a complex mix of the right state institutions, human capital and private firms (Porter 1990). Historically, African visions for IT-enabled services exports in the 1990s and early 2000s were fuelled by India’s and the Philippines’ successes and then shattered by a post-Financial-Crisis period in which the business process outsourcing (BPO) sector contracted. In North and South Africa, BPO sectors slowly matured, but across Sub-Saharan Africa, BPO firms from Ghana to Kenya either closed or turned to serving their small domestic economies. In some countries a surplus of talented youths emerged, who ended up exporting digital services via online labour platforms instead. 

Around 2016, three global trends began to impact the BPO sector: A dystopian literature on technological unemployment began to predict the imminent automation and re-shoring of most contact centre jobs in the sector (e.g., Ford 2015; Whadwha 2016); this coincided with new levels of protectionist rhetoric, driven by the UK’s Brexiteers and America’s Trump Campaign; and these two narratives, rhetorical as they were, overshadowed the third trend – a genuine uptick of the global economy in that year which led to real growth in the global BPO sector (including the growth of jobs for flesh-and-blood contact centre workers). 

As it became clear that BPOs were using the new technologies less for replacing their agents and more for turning them into tech-supported experts, the Covid-19 Pandemic crashed into the sector, fast-tracking the industry’s evolution towards (i) work-from-home (away from relying mainly on physical office work), (ii) spatial diversification (away from relying mainly on centres in a few Indian and Filipino cities), and (iii) gig platform work (away from relying mainly on fully employed workers). This brings us to 2022, where many BPO firms are actively looking to expand to new locations. In this context, it becomes interesting which combinations of ITES tasks and African locations are best suited to benefit from this window of opportunity.  

Tasks in IT-enabled services vary widely

The most sought-after and best paying jobs are those in which workers conduct complex tasks and hold great responsibility. These jobs will not be automated away anytime soon. Yet, to gain the abstract and tacit knowledge needed to perform such jobs competitively, workers require years of high-quality education and on-the-job training. These jobs tend to emerge and evolve in physical clusters of productivity. They come with high entry barriers for new locations and workers. On the other end of the task complexity spectrum are mundane jobs that are less fulfilling and less well-paid. These can be learnt much faster and in locations with lower availability of human capital. But in today’s fast-changing world of work, these are also the tasks most susceptible to automation. In IT-enabled services, the question for the most complex tasks, such as software development, is whether they can be learnt to be performed well enough to compete with programmers from more established clusters around the world. The question for the least complex tasks, such as data labelling, is whether the skills that workers acquire when conducting these tasks will lead them to slightly more complex work before their initial task is no longer in need of human workers. Where on this task complexity ladder – between the lowest rung of data labelling and the highest rung of software development – is the Goldilocks zone that holds the greatest impact for local development? That is a difficult but crucial question that needs to be explored by trial and error, and that will lead to different answers for different African locations.  

Hence, we tap into a wider debate on how to conduct industrial policy is concerned with how complex the tasks should be that a country pursues. Moon-shots often result in false starts, yet relying too much on comparative price advantages can lead countries to rely on dead-end sectors (Lin & Chang 2009). Which types of IT-enabled services tasks should aid-driven organizations like DSAA focus on? While some argue that the most complex IT tasks, such as software development, are the most fulfilling and future proof, others argue that in light of African demographics and factor endowments, it would be wiser to focus on less complex customer relations management or data labelling tasks to provide more jobs for less skilled youngsters who could then find their own ways up the career ladder. While labour intensive urban sectors have provided firms and workers with the tacit knowledge needed to climb the value ladder (Hausmann et al. 2013), it is not clear which digital services tasks actually help workers gain the types of skills needed amidst a digital revolution (Stephany 2021). Until this becomes clearer, it seems sensible to support a mix of tasks ranging from most to least complex and matching these tasks to the factor endowments of different African locations.

Locations for IT-enabled service clusters vary widely 

From the business perspective of for-profit global IT-enabled service firms, those new locations are best which can welcome the firm on a velvet carpet. A carpet carefully woven of stable governance, conducive regulations, an abundance of apt workers, a rich eco-system of other firms and training facilities, and technology and transportation infrastructure that makes it easy for the firm’s employees and clients to beam in and out virtually and physically. From the perspective of African governments and their development partners, however, it is clear that most of the yarn needed to weave such a carpet only becomes available once the firms that would like to land on such carpets are already in town.

 Hence, the study asks what makes a location more or less conducive for becoming the next IT-enabled service cluster. The right mix of governance, regulations, eco systems, apt talent, and infrastructure is needed. Globally aggregated country indicators are becoming ubiquitous, but if relied upon exclusively, these can be dreadfully misleading (see Annex 1 for my comments on a list of the most relevant indicators). A location’s two most important factor endowments are its human capital and its political stability.

To understand the human capital in locations where no BPO sector exists, some BPO firms and analysts have resorted to gauging the size of the local leisure and tourism sector (whose worker skill set arguably overlaps with that of contact centre workers). In this study, I argue that a better way of measuring IT-enabled services worker human capital is to examine the size and quality of the local pool of online platform workers. By this metric, Kenya clearly stands out on the continent, holding a large pool of BPO-related human capital that is mostly untapped by the BPO sector. 

To understand a country’s political stability, global indicators are less helpful. Countries can be stable and reliable democracies if the institutional framework is strong enough to facilitate peaceful political competition and curb the worst excesses of corruption. Or countries can be stable and reliable autocracies if the political leadership is strong enough to keep all factions in line and enlightened enough to steer the country in a developmental direction. Examining the six case countries, the study shows that Ghana is the best example of the former, while Rwanda is the best example of the latter. Ethiopia currently constitutes a case in which an attempt to switch from the latter to the former has failed. Kenya and Senegal are lesser versions of the competitive democratic Ghanaian model (with somewhat less peaceful political competition), while the Moroccan Monarchy is continually moving from soft authoritarian (secure enough to be tolerant of criticism) to hard authoritarian (increasingly fearful of descent). The study concludes that Ghana and Rwanda are politically most stable, but that all of the other case countries, besides Ethiopia, are also safe for investments. 

 Policy Advice – finding the Goldilocks zones

 The most crucial challenge is to lower entry barriers into the sector, given the demographic bulge in most African countries and the elitist nature of the digital services sector (most ITES jobs demand tertiary education from their job applicants). For this purpose, the three fastest growing lower-end IT-enable services tasks that could employ the highest numbers of African workers are customer relations management, social media content moderation, and data labelling. Hence, they are arguably the most important to pursue.  

Once a critical mass of workers has made it into the IT-enabled services sector, upward mobility becomes equally important. As a rule of thumb, the easier a task can be broken down into smaller pieces, the faster it can be automated away (Frey 2019). This means that the task with the lowest entry barriers – data labelling – may also be the task with the shortest life expectancy. Hence, while training higher-end digital services such as software development or data science holds fewer jobs, training such services is tremendously valuable for raising skill levels, regardless of whether or not these complex services can ever be exported. Higher technical skills may also become a prerequisite for the next generation of BPO worker applicants. But as much of Africa still trails other world regions in basic technical education and access to devices, first bringing down the entry barriers for the less complex ITES tasks is arguably the best strategy, leading to a greater pool of digitally literate youths who can then make use of further training programs to find their own ways up the task complexity ladder.

Geographically, strategies for boosting a country’s ITES sector can include more than one location. Most investments need to occur in the biggest city. But a handful of other cities are often even faster growing. Despite the strong pull of talent agglomeration to one national cluster, the flow of rural-urban migration forces farsighted governments to try to boost job growth in the bursting secondary cities. Some elements of the private IT-enabled services sector are surprisingly attuned to this and willing to take on the risk of locating away from the metropolis. These firms can be deliberately approached for partnerships, as they are taking on the unsavoury role of the first-mover to a new location, which should be viewed as a crucial form of impact sourcing, worthy of subsidies.

In their quest to boost the ITES sector across Africa, aid-driven organisations like DSAA can help private ITES firms by clearly pointing out what indicators show us which countries are stably governed. The study argues that, ideological preferences aside, systems both dominant (Rwanda) and competitive (Ghana) can be politically stable, and that both types have distinct advantages (e.g., orderly dominant systems can be better at planning and coordinating, while peacefully competitive systems can be better at fostering human capital development and creativity). This framework, applied comparatively, can give potential ITES investors some nuanced decision-making tools that can be difficult to glean from the political science trenches of the democracy-vs-autocracy debates.

The study concludes with more concrete advice for aid-driven organisations such as DSAA: (a) The types of companies to support and how (focus on large international firms that have a track record in job creation in the sector); (b) How to work with and support classical for-profit BPOs (help them assess and coordinate the environment in new locations); (c) How to approach the special case of social media content moderation (understand the great potential this task holds but be aware of its risks for workers); (d) What nationalities of the companies (try not to tie aid to firms from your country) (e) Organising ITES-sector support; and (f) How to best help BPOs expand into new African countries (harmonise ITES support activities with those of other donors working in the same space; build maps of knowledge by understanding the local eco-systems; then boost those eco-systems by supporting local private sector associations that are open to foreign direct investments in the sector; nourish the meso-level by involving potentially likeminded training institutions; once all this has been done, bring global firms to these locations and openly discuss the pros and cons of the geography but also make clear that longer-term development aid funding channels will remain open for worker training).

The study’s takeaway is that a window of opportunity seems to exist in 2022-23. If the above coordination efforts can be launched in time, they could spur the growth of IT-enabled services export clusters in several African cities.