Yes, it is true that African countries provide valuable data services for foreign digital platforms like Meta, Google, Amazon, and more. For example, in these foreign platforms, raw data from Africa is analysed and monetised (used for data brokering, bartering or targeting). Africans also provide data extraction and labelling services to these foreign platforms. Yet there is little local benefit to the data extracted from the continent. The term, Data Colonialism has been increasingly used to describe the unequal relationship that has emerged between global digital platforms (mostly in the west) and Africans.
Data colonialism, coined by sociology professors Nick Couldry and Ulises A. Mejias, refers to a new social order where the predatory extractive practices of historical colonialism are combined with the abstract quantification methods of computing. Simply, the idea is that, just as western countries unequally mined Africa for raw resources hundreds of years ago, they are doing the same today through data extraction from the African continent and its people without equal value.
To be frank, I am not entirely convinced of the theory of “Data Colonialism” as an explanation for the relationship between Africans and large foreign global digital platforms. I think the connection between colonialism and data extraction methods has yet to be properly highlighted in the literature. Regardless, data colonialism is not the focus of today’s post, though it is necessary to understand. The focus, today, is on Data Sovereignty and Localization policies, solutions that have been heralded as sufficient approaches to ending “data colonialism” in Africa.
Some examples of the unequal relationship between Global Digital Platforms and African data providers
Data colonialists scholars argue that Africans receive limited direct value from the vast amount of data extracted from the continent. This is a true fact. Africans provide value to foreign digital platforms but receive little back.
When scholars talk about the unequal value gained from extracting African data, they rarely provide actual material examples of what this value looks like. So, let us go through a few:
- Despite the fact that over 22 million users from Egypt alone use Instagram, META only has two (tiny) subsidiaries in Africa: one in Lagos, Nigeria and another in Johannesburg, South Africa. For reference, META has over 25 subsidiaries in Europe. In a similar vein, Google has 4 hubs on the African continent and one of these hubs is a developer’s space for google’s workshops and technical training programs. This limits the amount of jobs, as well as taxation that the countries may benefit from these platforms.
- Less than 1% of the world’s data center capacity is located in Africa. Neither Google nor META have a single data center campus in Africa, like the ones they have in Europe and the Americas. Generally, what these companies do for Africa is that they rely on a collocation model: they rent a space in an already established local data center to house their hardware. It is important to note that this “local” data center is usually still owned by foreign providers like Digital Realty.
- Lack of proactive regulation enforcement allows a lot of the foreign digital platforms to abuse Africans in spite of the large amounts of data they provide. As I have previously written in this blog, many African countries have comprehensive data protection laws that can rival any other country. Yet African workers continue to experience exploitation through off-shore gig work. The laws requiring consent for data analysis, extraction, labelling, etc are there. Laws for training against discrimination and setting standards for accuracy are present both in Europe and Africa. What is different is central enforcement power.
These are some of the realities that cause Africans to receive less value for their data. Data Sovereignty and Localization policies have been advanced as a solution to these problems.
About a month ago, I published a short piece interrogating the viability of Data Localisation policies in Nigeria. Read it to understand more about Data Localization.
Understanding Data Sovereignty and Data Localization
Data Sovereignty policies are laws that ensure that digital data is subject to governance structures of the country where it is collected. For example, In Nigeria’s Data Protection Act (NDPA), foreign companies like META/Google must ask Nigerians for consent to collect their data through cookies. On the other hand, Data Localization policies control the location of data. For example, requiring local data centers, and placing strict requirements on cross border transfer of personal or critical data for analysis.
Question: Can Data Sovereignty / Data Localization policies return value to Africa, providing solutions to the three material problems I highlighted above? I do not think so.
Data Sovereignty/localization laws vary in scope. Some merely impose rules on how to extract data from a country, others impose rules on where to store this data and who can gain access to it. Some of these policies are necessary for ensuring consumer autonomy and privacy. But others, especially those that require localization, often turn into burdensome and cost increasing measures for businesses. Coercive data localization or sovereignty laws can lead to capital flight. If these global digital platforms feel they are no longer receiving profits from conducting business in certain countries, they could leave. This is especially salient, considering countries that may not be considered to be providing critical revenue. The absence of these major global platforms undermines the consumer experience by denying African consumers the option of choice that comes with the free market.
Furthermore, data localization/sovereignty policies may contribute to monopoly capture of the digital platform industry. Adhering to multiple data policies on a country by country basis is expensive and requires large labor. It is those established firms that may have the capital wherewithal to comply with the laws. Thus, the laws serve as a barrier of entry to smaller firms that may have provided alternatives to these large global platforms who constitute the major harvesters of African data.
Some researchers have advocated for African countries to invest in viable domestic technological substitutes as a localised policy solution. However, advocates of these policies ignore the fact that African consumers have accrued network benefits on these apps like Instagram, and cannot easily leave them without social or financial consequences. Furthermore, investing in a domestic substitute is a huge capital investment that should not be in the periphery of any country. What is the opportunity cost? What part of the national budget will the country forgo in order to invest in platforms that can compete with Google and Meta? Furthermore, I personally, would not trust a platform that was developed through a partnership with the government or in which the government is a major stakeholder. Which leads me to my final point.
Data Sovereignty/localization laws often relegate a lot of power to the government. Some data sovereignty laws allow the government to arbitrarily inspect, control, and request access to data that has been collected from individuals in their jurisdiction. For example, in Nigeria (under the National Digital Economy Bill), the government often grants its own agencies “super-regulator” status. This can allow them to bypass the courts and demand “real-time” access to data for “national security.” Authoritarian governments can use data sovereignty/localization framing to justify surveillance of their citizens.
Considering my evaluation of Data Sovereignty/ Data Localization laws above, it is hard to see how these laws will fulfill their intended purpose of recovering the value of African data that is mined by foreign firms. For example, it seems to me that rather than creating more jobs these policies might even take jobs away. This is not to say that these data sovereignty/localization policies are completely bad. As was highlighted above, these policies can be useful in giving citizens autonomy over data privacy and holding private firms accountable. However, their benefits are certainly overstated. Still, all is not lost. There are other solutions to tackle the real issue of unequal value. In the next blog post we will discuss one of them which is derived from bargaining theory.