Towards a different kind of #DataMustFall? Data colonialism and inequality

In this series, we discuss data colonialism, a concept which is explored by LSE Professor Nick Couldry and Associate Professor Ulises Mejías of the State University of New York Oswego in their 2019 book, The Costs of Connection. In the first instalment, LSE PhD researcher Anri van der Spuy interviewed Professor Couldry to gain a basic understanding of the concept and how it relates to other arguments pertaining to capitalism in a networked age. In today’s instalment, she examines the potential practical implications of data colonialism for inequality.

Last week, we investigated the costs of moving our everyday lives online, in the context of ongoing coronavirus-related isolation measures. This week, we explore whether these costs of connection are borne equally: if processes of data colonialism expropriate human life itself to become direct inputs into economic production, is all data handled the same when and once it is extracted? In other words, are all data subjects created and treated equally under processes of data colonialism, or are some data subjects ‘more equal’ than others? To explore this question, we turn to an example in the global South before examining potential countermeasures.

The case of South Africa: from #DataMustFall to data for all?

An estimated 46% of South Africans are still offline today. While a plethora of barriers limit Internet use and adoption in the country, the affordability of data and the cost of devices are significant deterrents to digital inclusion. Compared to other African countries or even to benchmark countries in Latin America, the country’s mobile data prices are steep and have even been found to worsen over time.

In 2016, high data costs became the target of a popular social media campaign, #DataMustFall,[i] which was successful in gaining policymakers’ and regulators’ attention in the country. In his June 2019 State of the Nation Address, for instance, President Cyril Ramaphosa acknowledged the importance of bringing down data costs, explaining that wherever he had travelled in the country, people had ‘continuously raised the issue of the excessively high data costs’. Many also credit the campaign for the country’s Competition Commission data market inquiry in 2018/2019, which was initiated following ‘persistent concerns expressed by the public about the high level of data prices’. In an extensive outcome report published in December 2019, the Commission issued a raft of recommendations aimed at providing relief.

While some have criticised the Commission’s recommendations for being populist and going too far, others argue that the recommendations – which some mobile network operators have started implementing – do not go far enough towards achieving digital equality in the country.

The hidden costs of #DataMustFall

Because conditions online tend to echo conditions ‘offline’[ii], digital equality is likely to be some way off for a country understood to be one of the most unequal in the world. So while arguably a step in the right direction, the primary aim of #DataMustFall (i.e., lowering direct data costs) is therefore insufficient if digital equality is the objective. While focusing on demand-side challenges (e.g., ensuring people have the right skills to benefit from Internet access and use) would help to ensure more meaningful participation online, what about the less visible costs of digital inclusion?

Achieving digital equality in countries like South Africa arguably demands a more focused interrogation of the ideology of digital inclusion itself, including the hidden implications of the particular forms of connection introduced by processes like data colonialism. In their book, Couldry and Mejías warn that digital inclusion could come ‘at the cost of extensive data profiling, including practices that are predatory, discriminatory, exploitative, and simply degrading’.[iii] If this is the case, connecting more people to the Internet – whether as a result of lowering costs due to the success of campaigns like #DataMustFall, or some other measure – might inadvertently be exposing particularly vulnerable communities to costs for which they might be ill prepared.

Couldry feels strongly that the data colonialism argument is ‘at no point against basic connectivity’, but stresses that a broader picture of transformation is required. ‘Data works to discriminate,’ he points out, and ‘some people are far more dependent on and vulnerable to these changes than others.’ Online communities, in other words, will reproduce and even amplify offline inequalities because algorithms ‘can be engineered opaquely, unaccountably and therefore more effectively’ than ‘offline’ discrimination.

Because ‘colonialism’s sites of exploitation today include the very same West that historically imposed colonialism on the rest of the world’,[iv] inequalities will not only be amplified in global South contexts like South Africa, but will impact marginalised communities in the global North too (e.g., underserved communities like immigrants in high GDP contexts like the USA). And while only about a half of the world’s population are online, even people who are unconnected or sub-optimally connected could still be producing data fumes (even if unwittingly so) and are thus inadvertently being targets of processes of data colonialism.

Besides implications for the automation of inequality, investigating the less direct implications of #DataMustFall is important given South Africa’s policy interest in the Fourth Industrial Revolution (4IR)– the ostensibly inevitable ‘digital revolution’ which blurs ‘the lines between the physical, digital, and biological spheres’ and which is, unsurprisingly, stoked by data and its continuous expropriation. Despite it being highly problematic, enthusiasm for the 4IR is widespread in South Africa and beyond (e.g., its developmental potential has been touted by the African Union Commission).

Putting aside the apparent contradiction of punting a 4IR in contexts that have not even truly experienced a Third Industrial Revolution (i.e., basic Internet access), enthusiasm for a data-driven revolution is perhaps unsurprising given how commonly the problematic notion of ‘data as the new oil’ is touted by all and sundry. In The Costs of Connection, Couldry and Mejías explain that digital leaders in particular (e.g., China or the USA) tend to pressure countries in the global South ‘to open their digital borders in ways that benefit the social quantification sector’.[v] To do so, they’ve employed a raft of tactics ranging from discouraging data localisation laws to curtailing privacy protections; introducing substantial tax breaks to encourage digital giants to open local offices or operations; and other deregulatory conditions – sometimes in the guise of philanthropy (e.g., Facebook’s Free Basics) or the data provisions in free trade agreements.[vi]

But while ostensibly ‘free’, such ‘exchanges’ are everything but equal. With reference to other research on development and the trajectory of cross-border data flows, Couldry and Mejías warn that:

“The Global North still assumes the role of gatekeeper, as it did in the days of the telegraph and the telephone, and data flows continue to replicate the movement of resources from colony to metropolis.”

The Global North still assumes the role of gatekeeper, as it did in the days of the telegraph and the telephone, and data flows continue to replicate the movement of resources from colony to metropolis.

The global South, in other words, remain net exporters or sources of data (‘raw material’), while countries like China and the USA benefit from being net extractors of data. This is problematic for multiple reasons, including the fact that it leads to today’s ‘…colonizers becoming richer and more powerful, and the colonized underdeveloped and dependent’.[vii]

#DataMustFall, data dignity and/or data nationalism?

By promoting digital inclusion without simultaneously asking fundamental questions about the role of data (and data barons) within the wider frame of the new data colonialism, countries like South Africa risk falling further behind. Couldry argues that to better understand the ideology of digital inclusion, we might need to devise ways of rebalancing the scales so that processes of digital inclusion don’t just benefit a select few. Starting conversations about our relationships with data is crucial, but what of more drastic measures?

One suggestion, as Jaron Lanier has argued in a New York Times piece, is paying people for their data. But this so-called ‘digital dignity’ proposal introduces more questions than answers, the least of which being whether, if small families could earn ‘something like USD20,000 a year from data’ (as Lanier contends), the data of global South families will be worth the same as the data of small families in the global North. Couldry and Mejías both feel that instead of even tacitly condoning extractive data processes (as the data dignity approach would compel users to do), we should rather work towards questioning the normalisation and legitimisation of data extraction.

One way of doing so, Mejías has argued, is for countries (and especially those in the global South) to nationalise data as a public resource. We learn more about this data nationalism proposal in the final instalment of this series next week, when we talk to him about some of these questions and data colonialism’s implications for policy, development and inequality.

**This article was originally published by Media@LSE blog. It is published by Research ICT Africa with the author’s permission. It represents the views of the author and not the position of the Media@LSE blog, nor of the London School of Economics and Political Science or Research ICT Africa.

[i] Public campaigns in South Africa often feature the hashtag suffix #…MustFall, with other recent examples including #RhodesMustFall, to encourage the decolonisation of curriculums, and #FeesMustFall, to advocate for reduced university fees.
[ii] A binary approach to offline and online environments is generally not useful, but these differences still help frame indices concerned with equality and discrimination.
[iii] Couldry & Mejias, 2019: 68.
[iv] ibid:x.
[v] ibid, 105.
[vi] ibid, 105.
[vii] ibid, 104.