• March 20, 2023

Beyond The Hype: What You Really Need To Know About AI In 2023

By now, it’s probably clear to most people that artificial intelligence is going to have a fairly large impact on our lives. A few years ago, you might have been forgiven …

Google AI And Microsoft ChatGPT Are Not Biggest Security Risk, Warns Chess Legend Kasparov

Amid a flurry of Google and Microsoft generative AI releases last week during SXSW, Garry Kasparov, who is a chess grandmaster, Avast Ambassador and Chairman of the Human Rights Foundation, told …

What To Watch For From The Fed’s March Rate Decision

On March 22 the Federal Reserve will announce it update Fed Funds target. Markets suspect a 0.25 percentage point rise as most likely, with a slightly smaller chance of holding rates …

Policy experts and governments around the world are deeply dissatisfied with the way multinational corporations are taxed, especially big tech companies. The Organisation for Economic Co-operation and Development (OECD) has developed what it hopes will be a two-pronged solution: A global minimum corporate income tax and a tax based on where firms sell their products.

But some tax experts think even those broad reforms will fail in an increasingly digital global economy. As an alternative, they are proposing a tax on digital data. Not on its value or the income it generates, but on the volume of data.


For those of us used to a world of income or consumption taxes, it is a far-fetched idea. But administratively it would be not so different from a more familiar levy—a carbon tax that is imposed on the total volume of greenhouse gasses emitted by a firm.

Rethinking traditional tax concepts

This Thursday, my guest on the Tax Policy Center’s webcast The Prescription will be University of California Irvine law professor Omri Marian. He made the case for such a tax in a 2021 paper. World Bank economists Cristian Óliver Lucas-Mas and Raúl Félix Junquera-Varel have proposed their own version. University of Michigan law professor Reuven Avi-Yonah and co-authors suggested a similar idea earlier this year.

Corporations generally pay taxes on income generated from the sale of their goods or services (after deductions and credits for the cost of doing business). They may pay to the jurisdiction where they are located or where their product is sold.


But, says Marian, it is increasingly difficult to determine where income is earned, who owns the data, and what it is worth. What happens when, say, Facebook users receive a service in exchange for their personal information but no money changes hands?

He writes, “the data economy fundamentally changes the role of source, ownership, and monetary value…. In data-rich markets, it is not even clear [they] are theoretically meaningful constructs of legal-tax design. It is not clear that they assist, in any meaningful way, to identify one’s ability to pay.”

Taxing data, not income


In recent years, several countries tried to confront these challenges through licensing fees or excise taxes on telecom firms. Others tried a digital services tax (DST) on gross revenues from a tech firm’s sales in their jurisdictions. But these country-by-country DSTs run the risk of either double-taxing firms or missing taxable revenue.


The OECD’s Pillar 1 tries to confront these problems through a uniform worldwide system. Each country would tax income generated by marketing and distribution of digital products by a handful of large multinationals within its jurisdiction. Marian praises the intent of the OECD effort but believes it will fail.

In contrast, his tax base is the volume of raw data, no matter its use. The user of the data pays the tax. The levy could be imposed by countries or by sub-national jurisdictions such as US states.

Don’t try to anticipate how the data will be used, Marian says. Government never will keep up with those changes. Instead, simply measure the flow of data, and tax it. Tracking that activity, he says, is easy. Phone companies do it all the time.


Marian wouldn’t directly tax individuals every time they make a phone call or download a movie. Only heavy users would pay directly. That could make the tax easier to administer and more progressive.

Many variations

It does raise the question of who ultimately would pay the tax. Most economists believe corporate income taxes are paid by some mix of workers, shareholders, and other capital owners. By contrast, commodity or excise taxes (such as taxes on gasoline or beer and wine) are more likely to fall eventually on consumers.

Marian describes many data tax options: An excise tax something like the 1990s-era bit tax, a direct data tax, or a dividend tax on firms that collect and store significant quantities of personal information with a rebate to those whose data they gather.


Marian says a straight-forward data tax would work like this: New York State imposes a $1 tax on each gigabyte of data a firm collects from or transmits to New York IP addresses. It doesn’t matter from whom the data is collected, the type of data, or which machine uploads the data.

Avi-Yonah and coauthors would tax only downloads by for-profit firms, excluding small businesses and individuals. Lucas-Mas and Junquera-Varel favor a digital license tax and a levy on bandwidth.

Far more work must be done to design an actual tax and address administrative issues. And critics ask why governments would want to tax data, which, unlike, say carbon emissions, often has positive benefits. Some suggest a more familiar consumption tax could be a better alternative. But all these ideas are worth considering as technology continues to outrun the traditional corporate income tax.


Leave a Reply

Your email address will not be published.