2 The human rights turn to taxation

Taxation policies determine whether governments have sufficient fiscal space to realise the human rights of citizens and residents through the provision of social services and government programs. Whether taxation policies distribute burdens and privileges among different groups in society progressively or otherwise, they have crucial implications for material equality within nations. Internationally, taxation policies also influence whether needed revenues will flow to or away from less developed countries, thus impacting equality among nations.Footnote ⁴ The efforts of groups like the Tax Justice Network in scrutinising tax cuts for the superrich, closing international tax havens, promoting financial transparency (automatic bank information-sharing and registration of beneficial ownership), unitary taxation and the UN Tax Convention provide illustration of some of the work that is being done to bring a human rights lens to international taxation policies. We are seeing some UN human rights bodies (eg, UN Committee on the Rights of the Child which investigated Ireland’s tax policies’ facilitation of profit shifting) involved in the push for tax reforms.Footnote ⁵

Important proposals to revitalise social and economic rights require funding and often entail taxation in some form. Chief of these proposals is the expansion of the systems of social protection which is a recognised human right under Article 22 of the Universal Declaration of Human Rights (UDHR) through the establishment of universal basic income (UBI). As United Nations Special Rapporteur for Human Rights and Extreme Poverty, Philip Alston argued that in the face of increasing job insecurity as a result of globalisation, the rights to social security and an adequate standard of living are best secured through the establishment of UBI.Footnote ⁶ His argument provides an explicitly human rights grounding for UBI, a topic that has attracted renewed discussion. However, UBI remains a politically controversial proposition and disagreement exists over its human rights implications. Some of the more compelling critiques of UBI point out that a UBI that provided social security and an adequate standard of living for everyone has to be substantial in amount; however, a substantial UBI will be so expensive as to entail negative human rights consequences. Human rights lawyer and economist Philip Harvey argued, for example, that a substantial UBI had an ‘opportunity cost’ in that its implementation would preclude the provision of other social services needed to fully realise other economic and social rights recognised under the UDHR.Footnote ⁷

While Fasan does not engage in a critique of UBI, she has argued for a more modest and targeted use for ‘data capital tax’, namely, to fund underfunded medical research. This suggestion is meant to show that DT could enable the state to support neglected medical research that benefited people’s health. A more commonly suggested alternative to UBI is a job guarantee for those able to work (coupled with targeted social protection for those unable to work). It is argued that a job guarantee is a more realistic, lower-cost alternative to UBI that moreover addressed challenges to the right to work in the context of increased automation.Footnote ⁸ Regardless of where advocates stand on the UBI versus the cheaper alternatives debate, it is clear that the existence and sufficiency of funding sources is a major consideration in the justification of the options that should be pursued.

Clearly, new taxes that allow for ambitious government programs should interest human rights advocates. For example, the expansion of social protection, which is crucial post-COVID, could be funded from new taxes including DT.Footnote ⁹ DT is an example of a number of proposed tax reforms which problematise the inadequacy of existing taxation policies for our age.Footnote ¹⁰ They theorise how new wealth is being created in a data-rich economy but is escaping taxation or redistribution to benefit ordinary people.

3 Sharing wealth while protecting from harm: some considerations for DT

The new context of the digital economy, however, not only provides new opportunities for generating funding for promoting social solidarity and people’s welfare but also demands protections against new harms such as those entailed by surveillance capitalism.Footnote ¹¹ Indeed, among the mainstream of the human rights movement, the need for protection is probably more appreciated than the need for wealth sharing.Footnote ¹² However, as we have seen in the foregoing discussion about the opportunity cost of UBI, a proposed course of action may advance human rights in one sense (eg, expand the social safety net) but may still have a suboptimal overall or net impact on human rights. A course of action that produces optimal overall human rights outcome should create synergies rather than tensions between the achievement of various human rights.

Thus, taxation that aimed towards benefit or wealth-sharing in the digital economy by expanding the safety net or through other means can be harmful to human rights overall if it contributed to entrenching harms to other human rights. As an example, expanding the safety net through UBI may well provide relief to those displaced by automation but in the same breath it may also become a palliative that reconciles society to unrestrained profit-seeking in which technology is used to make jobs more exploitative or to cause the disappearance of employment relations regulated by democratic labour law.Footnote ¹³ In this scenario, UBI could neutralise calls for greater market regulation and labour law reforms, contributing to the entrenchment of rather than the challenge of the status quo, and therefore may be detrimental to human rights in the long term.Footnote ¹⁴ However, such critiques of UBI are not insurmountable. They do not deny that UBI helps, but they do imply that its justification and limitations need to be made clear. They demand that UBI should be reframed so that it accompanies and strengthens rather than competes with other reforms aimed at challenging unequal and exploitative relations.Footnote ¹⁵

To see how DT is superior to other proposals for sharing the wealth of the digital economy, consider data dividends which have been one popularly proposed ways of capturing wealth to benefit individuals. Like DT, data dividends presume that data is valuable; proponents of data dividends propose that data’s value should then be apportioned to individuals on the basis that they contributed to value creation. California Governor Gavin Newsom, for example, had originally suggested that the California privacy law implied payment of data dividends to Californians, echoing similar ideas that social media companies, for example, should pay users for using their data.Footnote ¹⁶ Putting aside the question of administrative complexity or practical impossibility of computing the value of individual data,Footnote ¹⁷ taxing big data (to fund social services among others) was preferable to data dividends because while both entailed benefit or wealth sharing, DT avoided harm to human rights where data dividends would likely entail harm. This was because assigning monetary value to individual personal data may also enable corporations to assert that individuals’ privacy rights to such data has been sold or surrendered to them. Surely, the right to privacy means more than creating data dividends for those who can successfully negotiate with corporations to obtain them; the protection of privacy in its dignitarian meaningFootnote ¹⁸ seems in tension with its commoditisation. Unlike data dividends, DT promises to capture wealth from data for redistribution to ordinary people without being entangled in such complications over the meaning and future of the right to privacy.

To the contrary, DT has Pigouvian potential. Like sin taxes on alcohol and tobacco, for example, taxing big data could be regarded, at least in theory, as a discouragement of excessive collection, storage and processing of data, and all the attendant harms to privacy and other human rights that excessive datafication entails. Indeed, if DT functioned to discourage excessive datafication, whether DT funded UBI, social services or a specific public good, DT would have contributed to challenge surveillance capitalism, not entrench it. This does not mean discouraging excessive datafication can or should only come in the form of taxing data; indeed, other reforms such as stronger privacy protections and more democratic mechanisms for governing data may be needed to achieve that end. Whether DT actually discouraged the harms to human rights emanating from excessive datafication may depend on the existence of other protective mechanisms which complement it in the same manner that UBI needs to be complemented with other reforms.Footnote ¹⁹

4 Exacting rent on data and technology: DT as a technological gains tax

Where RtS and some theories of DT appear to differ is in their emphasis of where value emanates. RtS refers to ‘benefits’ that arise from ‘scientific progress’ and the ‘application of science’ (ie, technology), whereas some theories of DT emphasise that in a digital economy, value largely derives from big data. Thus, these theories of DT assume that what needed to be captured and distributed was the value of big data. In Omri Marian’s theory, the value of big data corresponded with its volume not necessarily its realised monetary value; ie, corporations should be taxed based on the volume of data they collected, stored and processed rather than the monetary value they realised from big data.Footnote ²⁰ This recognised that data was the new currency or tax base. According to Marian, in a data-rich market economy, the ability to collect, store and process increasingly voluminous data was the best proxy for ‘ability to pay’, indeed better than monetary income itself.Footnote ²¹ Taxation policies that insisted on taxing monetary income rather than data ignored the impact of big data on value creation and the markedly different abilities of corporations to collect, store and process them, thereby allowing corporations that had the best ability to pay to remain insufficiently taxed. While tax reforms such as the digital services tax may capture some of the monetary income realised from digital services, eg, online advertising, such reforms will remain inadequate until data itself rather than merely realised income is taxed.Footnote ²²

In this symposium issue, Vipra supports the emphasis on big data and justifies largely ignoring other factors such as algorithms and computing power which are private properties of corporations. Her concern is the pauperisation that arises from the inability of masses of people to sell their intellectual labour as a result of the rise of AI and her proposal is to counter this through UBI funded by DT. Big data was, Vipra argues, a proper subject of ‘rent’ ie, taxation in the form of DT. Vipra’s notion of ‘data rent’ is fashioned from Thomas Paine’s theory of ‘ground rent’. In Paine’s Agrarian Justice, he theorised the justice of exacting a tax (ground rent) from landowners and distributing it as a payment to all (a form of UBI) in recognition of the fact that land, though now privately owned, belonged to and should benefit everyone in common as it used to be before enclosure.Footnote ²³ Vipra’s notion of data rent rehearses Paine’s solution but in the context of the rise of AI greatly devaluing intellectual labour and thus preventing masses of people from making a living from such intellectual labour. The theory relies on the argued correspondence between data and ‘economically relevant intellectual capacity’. Similar to Vipra, other proponents of UBI as a solution to AI-induced job displacement also theorise that the value created by AI partly belongs to the masses of people given that their data have contributed to the big data necessary to evolve AI capacities.Footnote ²⁴

However, the assertion that value arises from big data clearly involves a conceptual shortcut. Data themselves have no value unless converted into things like insights, predictions, and data products which are of benefit to someone, whether this be realised as monetary value or not. As corporations sell these predictions and data products to advertisers and the behavioural futures markets, corporations realise the value of data as monetary gain; but corporations may well use insights, etc. in other ways to benefit themselves through greater operational efficiency or what not. States, communities, non-profits or others too may as well use insights, predictions, and data products derived from big data for their own or the common good without these things being any monetary value. The point is that data is not inherently valuable but is imbued with value through being related to other data through the application of advanced analytic techniques. Big data does not analyse itself after all; you crucially need data scientists to convert data to the kind of value that can be of benefit to someone. For big data to have value, all the needed analytic techniques, sociological knowledge, technological infrastructure, etc. must combine and be applied to convert data into value.

In contrast with Vipra, I argue that other factors that are combined with big data, ie, algorithms and computing power are also proper subjects of rent. This is where RtS comes into play; we also need to pay attention to data science and infrastructure rather than merely data. Scientific ideas and discoveries (the labours of generations of scientists) underpin technological progress; this is as much true of digital technologies as other technologies. RtS underscores that these scientific ideas are the heritage of humanity and problematises the privatisation of its benefits. More concretely, as Mazucatto has underscored, states have contributed to technological innovations in the form of, among others, public investment specially in the early high-risk stage of research that led to technological innovations.Footnote ²⁵ Apple’s iPhone thus could not have been possible without the component technologies (the internet, the microchip, the touch screen, etc.) which were developed with substantial public investment. If, as Paine and Vipra have suggested, the exaction of rent from corporations which privatised what is supposedly common property was important to restore social justice, then we had been missing mechanisms for the recovery of rent on the scientific heritage as well. We have also been missing mechanisms for ‘pre-distribution’ by which part of the gains from technological innovations that have been commercialised should return to the public.Footnote ²⁶ Failing to capture rent on scientific and technological progress was socially unjust because it meant a few corporations undeservedly benefited from the labours of generations of scientists, and public funding by the state, while ordinary people were excluded from what is supposed to be a common inheritance.

Thus, I argue that DT could be thought of not only as ‘data rent’ but also as rent on data and technology. Big data need not be strictly differentiated from other factors (algorithms and computational power) with which it is combined to create value. Even though, as Marian proposed, for administrative simplicity, the computation of DT itself could be based only on the volume of data and no other things, DT taxed the ability to collect, store and process increasingly voluminous data which depended on technological factors. From this light, DT functioned to introduce a kind of technological gains tax with respect to data science and its application in monetising or deriving economic value from big data, a measure that had been missing from previous eras of technological progress.

5 International dimensions and feasibility of DT

So far, we have discussed DT at the level of the state which could impose it on corporations to benefit citizens and residents in some way and protect them from harm. Especially when DT is suggested to fund the expansion of social protection such as UBI, we have imagined the decision-making and implementation to take place at the level of the taxing state. However, if the elaboration and implementation of DT were to lead to an overall positive human rights outcome, the impact of the implied taxation policies on international equality should also be of concern. As mentioned, internationally, taxation policies influence whether needed revenues will flow to or away from less developed countries, thus impacting equality among nations. If we were to seriously insist that the benefit of big data should belong to everyone, then DT should be equitably shared internationally.

Marian has highlighted that states could easily enforce DT within their territorial jurisdictions given that data flows through some physical infrastructure which states can effectively control.Footnote ²⁷ This might be a practical advantage that makes DT attractive to states desirous to raise revenues without waiting for an international agreement to be formed or the need for coordinating with other states. Christians and Magalhães (in this symposium issue) also argue that redoing digital services tax as income tax by expanding the withholding tax system to include certain digital service fees would enable states to tax tech companies without coordination with other states including, importantly, the United States. But if DT were collected where the infrastructure for the flows of big data exists and benefited only those countries, then, this raises concern about the justice of not sharing the value of big data with everyone. As Brauner argues, if countries benefit in proportion to how much data flows through the infrastructure they host in their territories, the jurisdictions with the most developed infrastructure for big data will likely benefit the most.Footnote ²⁸ Moreover, not needing international coordination, data taxes enforced unilaterally like the digital services tax will have the side effect that nations will compete to attract corporations through more lenient tax policies rather than unite over their more just taxation.Footnote ²⁹

It is thus important to consider the feasibility of DT not only as a tax that can be administered and collected domestically, challenging and important as that question may already be, but also as a tax that is regulated by international rules that ensure it is equitably shared across developed and developing countries alike. More familiar concerns with income taxation of multinational corporations cast a shadow on the question of DT’s feasibility as a tax with international implications. The main troubles with business taxes lie in the ability of multinationals to shift their profits to tax havens, and the reaction of states, acting independently of other states, to counter the erosion of their own tax bases by lowering their tax rates and/or introducing measures like financial secrecy that exacerbate the problem. Both problems result from decisions made by states. Hugo and Løvold have emphasised that harmful tax competition is a cookie-cutter instance of an ‘international cooperation problem where individual rationality is at odds with collective rationality’.Footnote ³⁰ Thus, the feasibility question is largely a question of how to convince states to work together and share taxes in a fair manner.

On the bright side, states are not unfamiliar with the notion that tax competition is harmful to everyone in the long run and that it must be stemmed through international cooperation and the development of international law. A century of international tax treaty-making attests to this. This includes the OECD’s success in forging the two most significant multilateral tax agreements and its current efforts to facilitate international negotiations on solutions to the tax challenges of the digital economy.Footnote ³¹ In the EU, this history also includes the 1967 decision to create an EU-wide system of Value Added Tax (VAT) as part of the programme to create a single market and the non-binding 1997 Code of Conduct on business tax that encouraged non-competition in tax matters.Footnote ³² Hugo and Løvold, furthermore, note that today the 20 largest economies support the so-called ‘ABC of tax transparency’ originally proposed by the Tax Justice Network, namely, automatic exchange of financial information, beneficial ownership transparency, and country-by-country reporting, which build towards a system of unitary taxation.Footnote ³³

The new watershed moment for international cooperation on this matter is the recognition by many non-European and developing countries that solutions need to be even more globalised than what the OECD is already currently pioneering.Footnote ³⁴ A key idea championed by advocates of the UN tax convention is unitary taxation whereby every state is enabled to tax multinational groups on their global income in proportion to that country’s share in the multinational groups’ workers and sales.Footnote ³⁵ International fairness is operationalised by seeing the tax base as global (the multinationals’ global income) as well as having a formula for sharing it that is accepted as fair across developed and developing countries.

The foregoing reflection implies that DT proponents not only must but can aspire towards global cooperation; anything less will fall short.Footnote ³⁶ Thus, if DT were established within the EU, that might signify a level of international cooperation that improved fairness within the EU. But that would still attract the same criticisms as those levelled against the OECD’s proposals on the taxation of multinationals, namely, that the net impact on developing countries is negative (See, Oxfam 2021; McCarthy 2022). Ensuring international fairness in DT could learn from the current proposal for a UN tax convention. That is, states need to imagine the tax base as global (and as I have argued, the right to science helps in this regard). Moreover, a fair formula for sharing DT across jurisdictions must be found.