Picture: 123RF/SOLARSEVEN
Picture: 123RF/SOLARSEVEN

Covid-19 has demonstrated how data can provide new sources of evidence for public policy-making. Internet search histories are being used to detect unknown outbreaks and in many countries governments are using cellphone location data to track population movements and contain the spread of the virus.

The pandemic has also reminded us that data “ownership” is heavily skewed in favour of the private sector. Capturing, organising and analysing data is a multitrillion-dollar business and the most valuable data — private (often high-frequency and geolocated) data — belongs to banks, mobile network operators (MNOs), and tech giants such as Facebook, Amazon and Google/Alphabet.

In economic terms data is similar to roads, parks and clean air. First, it has high upfront costs and low marginal costs. It’s expensive to set up a data collection system, but once it’s in place it’s cheap to collect the data. Second, data is “non-rivalrous”. It can be provided to additional users at little or even zero cost and it is not depleted as more people use it. This differs from other resources, such as oil or gas, where the use of one person prevents the use of the same oil or gas by another.

Finally, data has “externalities”. The use of data by one person affects the costs and benefits of others. We tend to focus on the negative externalities of data, as when private firms harvest our data and sell to third parties without our knowledge or consent. The Facebook-Cambridge Analytica data scandal is a case in point. We forget that data also has positive externalities. When different data sets are combined, new knowledge is produced and new value is created.

From a public policy perspective this suggests we should treat data the same way we treat roads, parks, and clean air: as a public good. To an extent we already do that. Most academic research is government-funded and national statistical services, such as Stats SA, collect and publish official socioeconomic and demographic data. However, the largest and fastest-growing repositories of data are held by private firms.

This gives them market advantage and allows them to expand beyond their core businesses. Google/Alphabet, through its Waymo subsidiary, is the market leader for autonomous vehicles, not because it knows how to build cars but because it has access to more location data than anyone else. Similarly, mobile network operators such as MTN and Vodacom, are making billions on banking and insurance, drawing on their knowledge about mobile phone users.

The concentration of data in the hands of private companies slows down innovation and growth. When a few dozen companies have access to the most valuable data only a few dozen companies can contribute to inventing new things. A recent report by the Organisation of Economic Co-operation and Development suggests that if data were more widely shared many countries could enjoy gains worth 1%-2.5% of GDP.

Ironically, the technology and infrastructure upon which many of these companies base their success were developed with taxpayer money. Google’s algorithms were developed with funding from the US National Science Foundation. The internet, touchscreen displays and GPS came out of the US defence department. It seems fair that the data they collect is made accessible to all of us.

Companies are already moving in the right direction. In April Microsoft launched an open data campaign and announced plans to create 20 data-sharing groups by 2022. Similar steps have been taken by Spanish bank BBVA and British pharmaceutical company GlaxoSmithKline. However, they are the exception rather than the rule. Most companies still treat data as private property.

So what needs to be done? First, we must acknowledge that market forces alone will not solve the problem. Too much data is already concentrated in the hands of the private sector, and government intervention is needed if we are to benefit — collectively — from the data revolution.

In concrete terms, this means governments must adopt comprehensive policies that go beyond the protection of private information and promote equal access to data. No-one should be granted exclusive rights to public-sector data. A recent report from the Bennett Institute of Public Policy at Cambridge University even suggests private companies should be obliged to make their data accessible to third parties and ensure interoperability between platforms. The reduced incentive to invest in the data economy could be offset by issuing time-restricted licences similar to patents.

Privacy concerns will also have to be dealt with in more innovative ways. Restricting access and aggregating data are not the solution. Computation and encryption-based alternatives should be adopted. National statistical services could play a big role, which means their mandate should be expanded. In addition to their core functions, they could even serve as custodians of private data collected by others.

Covid-19 has made it evident that most governments are behind the curve. Their policies are inadequate and they do not have the mechanisms in place to systematically incorporate data in their policy-making processes. While the private sector has generally responded responsibly to the pandemic by sharing data and offering analytical capacity, it is unsustainable that our collective interests are left at the mercy of profit-driven firms. Policymakers need to act, and it is urgent.

• Lynge is a senior lecturer at the Wits School of Governance.

subscribe

Would you like to comment on this article or view other readers' comments?
Register (it’s quick and free) or sign in now.

Speech Bubbles

Please read our Comment Policy before commenting.