Data is the new oil. This now-ubiquitous adage is a reflection of the immense monetary value generated by the literally quintillions of bits and bytes transferred over electronic wires. Many companies that rely on data for their money-making activities are doing very well indeed—the combined revenue of Google and Facebook, among the world’s biggest data imbibers, was a staggering US$192bn in 2018, according to Statista.
The outsized value of data in today’s economic landscape is raising new questions about how governments can grab a share of the bounty—and, by extension, return it to the public sphere. Some ideas, such as the digital dividend currently being explored in California, include the possibility of a tax on large data-collecting firms, which would then redistribute the money raised back to residents in the form of regular payments.
The idea of placing a levy on data as if it were any other asset on a firm’s balance sheet, such as an office building subject to property tax, runs into a number of practical problems. For starters, many companies have no idea how much their own data is worth, or even how to place a value on it. This problem has been subject to a variety of analyses, with little consensus in sight. Some believe it to be largely intractable. “Usually it’s the forging of linkages between different data points that creates value, rather than data in their pure form,” says an executive at a San Francisco Bay Area-based software company. He notes that different companies may value the exact same data in different ways. In addition, data is often only valuable in aggregate, rather than as a single slice.
Can you spare a byte?
Still, the debate around how to more equitably distribute the spoils of the so-called data economy is gathering steam. In October 2019 the OECD, a club of mostly rich countries, released a draft of a new global tax regime for companies operating in the digital space, many of which make money by offering free online services—like search or social networking—in exchange for user data. These data are then used as the basis for highly targeted advertising, the ultimate source of these firms’ income. Although the OECD framework does not assign a one-to-one equivalence of “data for money”, it specifically identifies data as an asset collected by large, globe-spanning organisations which are often physically far removed from the users who generate it.
Under the new proposal, taxes would be imposed where the data itself are collected and revenue generated, rather than in a third country. If ultimately adopted, it will likely pre-empt unilateral taxes imposed by individual countries, such as a French levy that has been decried by American companies and the US government alike. “Some countries are saying, ‘there is value being created here—we should be able to tax that’,” says Marijn Verhoeven, lead economist in the tax division at the World Bank. “[They’re saying that] it’s not appropriate for these platform companies to post this profit in a low-tax jurisdiction.”
The OECD framework does not stipulate how the resulting tax revenue should be distributed, but at a macro-level the aim is to even out the imbalances of global cash flows generated by large border-spanning firms. “You have data that you collect in a country where you may not even be present,” says Mr Verhoeven. “There’s no one to tax in the country itself.”
Governments’ ability to unlock value from the data sloshing around the private sector may extend further yet. Nikos Askitas, co-ordinator of data and technology at the International Data Service Center under Germany’s Institute of Labor Economics, says governments should consider compelling firms to pay a data tax in data itself, rather than (or in addition to) a traditional monetary levy. Such a tax would apply to two types of companies: the first, those such as Amazon who are so large and structurally significant that the data they hold can reveal useful information about markets or the broader economy to policymakers. “Data has value for the big picture,” says Mr Askitas. “What you get is timely intelligence [for policymaking], which in many cases has no monetary equivalent.”
The second group is comprised of companies that “take payment” in data and rely on advertising for their revenue. These companies—especially when they grow so large that they dominate their field, as Google and Facebook do—hold proprietary information on their users via the data they collect. According to Mr Askitas, compelling them to hand over these data (in an aggregated, anonymised manner to protect consumer privacy), and then placing it in the public domain or in the hands of academic researchers could help spur competition and shine a light on the inner-workings of the economy itself. Mr Askitas cites as an example Google Trends, which uses its data pool to summarise the activities of its vast user base and allows the public to discover information on a variety of topics, such as the input frequency of popular terms on a local, national or global scale.
Transparency international
Such data would need to be granular enough to provide visibility of the firm’s data-collection efforts without necessarily revealing company secrets, says Mr Askitas. The entire scheme would need careful oversight to counter the risk of its abuse for political gain—think Donald Trump threatening Amazon because its CEO, Jeff Bezos, owns the Washington Post. It would also need international calibration, so that data gathered in one jurisdiction is not gratuitously piped into another government’s coffers simply because the company that gathers it is based overseas. Yet as the political debate over the role of big tech builds momentum, the idea of governments using private-sector data to level the playing field—either via monetary means or otherwise—will gain a greater following.
