1. Introduction
By and large, digital constitutionalism – roughly understood as the constellation of scholarly and policy discourses exploring the relationship between constitutionalism and the socio-legal challenges of the digital revolutionFootnote 1 – has not focused on value creation, extraction, and distribution of data/informational capitalism.Footnote 2 To be sure, legal scholarship has long investigated the relationship between the digital revolution – with its underlying political economy and legal infrastructure – and social justice.Footnote 3
However, authors rarely thematise these issues within explicitly constitutionalist frameworks.Footnote 4 Scholars using such frameworks, on the other hand, mainly focus on questions such as access to the Internet; free speech and disinformation; privacy and data protection; procedural guarantees (transparency, participation, fairness);Footnote 5 digital administrationFootnote 6 and justice;Footnote 7 applicability of constitutional law standards to private actors via (some variation of) the ‘horizontal effect’ of fundamental rights.Footnote 8 Moreover, most constitutional analyses take the regulatory capacity of modern states somehow for granted and do not account for tendencies towards the re-feudalisation of socio-political relationships deriving from the business models of data economy.Footnote 9 Put differently, the impact of digital and algorithmic innovation on societal processes and on (the legitimation of) both national and non-national political institutionsFootnote 10 is hardly linked to the latter’s capacity to effectively pursue policies such as social justice and egalitarian objectives.Footnote 11 At an even deeper level, constitutional lawyers do not thematise the dangers coming from the excessive datafication of societyFootnote 12 as a self-standing problem.
Against this background, this Article introduces a symposium that contributes to filling this gap, using data taxation as an entry point. More specifically, it sketches the elements of a research and policy agenda concerning data taxation within a (digital) constitutionalist framework. While the single contributions adopt distinct approaches and advance different proposals, they share the view that the externalities of informational capitalism are issues of constitutional significance.
Starting from this view, this Article and the symposium (try to) keep four macro-issues together: (1) the impact of excessive datafication on contemporary societies; (2) the role of data in contemporary economy and the justifications to its taxation; (3) concrete tax design, coherent with the regulatory purposes of a (digital) constitutionalist agenda; (4) the interaction of data taxation regimes with other legal regimes and social justice issues, also at the global level. The interlinked questions underlying these issues constitute the building blocks of an expanded digital constitutionalism, which includes the socioeconomic dimension not exceptionally but structurally.Footnote 13
Put differently, this Article – and the entire symposium – takes on the issue of data taxation to explore the interconnection among such questions, both analytically and normatively.Footnote 14 The first-order goal is to increase the dialogue among different strands of scholarship – constitutional law, law and technology studies, ‘law & political economy’ (LPE), and tax law, among others – that do not necessarily speak the same language. The second-order goal is to expand and strengthen the analytical and normative scope of digital constitutionalism.
These goals are based on a specific view of the normative purposes of digital constitutionalism: if the latter aims to rise at the level of complexity required by the challenges posed by the digital revolution and informational capitalism, it needs to address issues related to value creation, extraction, and distribution, also through interactions with legal fields that do not speak the language of constitutional law. If digital constitutionalism aspires to be authentically normative, it cannot address such issues as contingent, accidental elements.Footnote 15 To be fully normative, digital constitutionalism needs to be (also) an economic constitutionalism.Footnote 16
After this introduction, the Article proceeds as follows. Section 2 focuses on the link between the digital revolution and (the legitimacy of) constitutional states, notably when it comes to the latter’s role in value creation, extraction, and distribution. Section 3 identifies such an issue as a gap in digital constitutionalism and opens the way to the following sections. Section 4 is divided into four subsections. Section 4.A highlights that, to address the issues mentioned coherently, one has to start from a critical approach to datafication. This means that excessive datafication needs to be seen as a source of problems in itself, to be addressed as an autonomous object of regulation. Section 4.B highlights the role of data as an economic factor within the data economy, beyond its monetary value and the specific conceptualisations in different legal fields. Analysing the role of data in the contemporary economy, one can individuate normative justifications for its taxation, even in the light of constitutional limits to states’ taxing power. Section 4.C moves to tax design issues, highlighting the need to include Pigouvian, progressive, and rent-targeting elements. Finally, Section 4.D puts these issues within the context of economic governance, highlighting the role of (global) institutions in creating, extracting, and distributing economic value, and the political nature of the underlying policy choices. Section 5 concludes.
2. Böckenförde’s dilemma, governability and the (digital) constitutional state
‘The liberal secularised state lives by prerequisites which it cannot guarantee itself.’Footnote 17 Böckenförde’s (in)famous line looks at any lawyer standing at the gates of constitutional theory. At its core, it is a sobering reminder of the insufficiency of modern (constitutional) law in generating the preconditions for its own normativity. When god(s) cannot be invoked anymore – not directly, at least – and nationalism has generated monsters, when justice and values mean something different to each social group, the legitimacy of political institutions in post-war constitutional states has come to rely on their capacity to adopt collectively binding decisionsFootnote 18 while simultaneously preserving room for conflicts to emerge and mediating them.Footnote 19 Post-war liberal, constitutional states with capitalist modes of production are thus characterised by an intrinsically precarious balance which puts their own ‘governability’ into question.Footnote 20
Indeed, the capacity of political institutions – both national and supranational – to generate, extract, and distribute value, while also striking balances among societal actors, is crucial to their legitimation and, ultimately, to the performance of their societal functions.Footnote 21 Effective, socially legitimated systems of value creation, extraction, and distribution are not only instruments in the hands of political apparatuses aimed at preserving social peace. At a deeper level, they are essential to the capacity of political institutions to adopt consensus-based, collectively binding decisions, that is, to the performance of their societal functions.Footnote 22 As instruments of social justice, tax systems are crucial to both socioeconomic governanceFootnote 23 and the legitimation of modern political authority.Footnote 24
Such elements are relevant to digital constitutionalism. The digital revolution, the platformisation of socioeconomic relations, and data- and automation-driven business models have profoundly impacted the political economy and the public sphere(s) underlying contemporary societies.Footnote 25 The COVID-19 pandemic has accelerated pre-existing dynamics, touching upon the relationship between authority and freedom, state and society, politics and economy, collective and individual actors.Footnote 26 The public sphere(s) where individuals and groups generate debate, contestation, and conflict have become extremely different from those presupposed at the foundational time of modern constitutionalismFootnote 27 and often go beyond the territorial borders of nation-states. These developments, in turn, take place alongside processes of political-economic globalisation, dispersion, and fragmentation triggered or dominated by neoliberal policies. In this context, the value of traditional procedures legitimising constitutional states – especially elections and other institutions of representative democracy – is more and more eroded.Footnote 28
These developments are further linked to the crisis of economic governance models presupposed by post-war constitutional states. In addition to the race to the bottom triggered by the competitive alignment of regulatory, fiscal, welfare, and labour protection systems,Footnote 29 and by so-called austerity policies, the business model of the data economy impacts socio-political integration by also affecting the capacity of value creation and distribution – that is, taxation and welfare systems – presupposed by modern constitutional states.Footnote 30
Models of value creation based on digital services and finance, personalised advertisement, big data analytics, cryptocurrencies, smart contracts, metaverse(s) – in one phrase, informational capitalismFootnote 31 – accelerated and exacerbated pre-existing social dynamics of late-20th-century economic globalisation, weakening the capacity of political institutions to govern social processes also through economic distribution. Following processes of globalisation, dispersion, and intangible-isation,Footnote 32 the data economy makes traditional taxation systems – especially the income tax – much less effectiveFootnote 33 and aggravates the fiscal crisis and ‘base erosion’ that emerged even before the global neoliberal turn of the 1980s.Footnote 34 Informational capitalism does not influence only individual behaviour and how societies collectively reach (presumptive) consensus on specific issues and accept political authority as legitimate. It also affects political institutions’ capacity to govern the economy via collectively binding decisions and politically legitimated law-making. The externalities of profit-driven datafication affect the ability of politics to produce socially legitimised decisions;Footnote 35 of science to produce socially shared truth;Footnote 36 of medicine to improve collective health;Footnote 37 of economy to create value for the whole society.Footnote 38
3. A gap in digital constitutionalism
By now, there is a vast literature on how law ‘codes’ data, making it one of the factors of production within the data economy/informational capitalismFootnote 39 and contributing to ‘datafying’ other factors of production.Footnote 40 Social justice issues have been explored especially by labour law scholars, for example in the context of platform work,Footnote 41 workplace democracy,Footnote 42 or the impact of digital technologies on processes of value-creation and -distribution at the global level.Footnote 43 Tax law scholars, for their part, investigate how informational capitalism affects states’ fiscal capacities, potentially contributes to tax avoidance,Footnote 44 or how Artificial Intelligence (hereafter AI) and digital technologies may optimise tax systems, making them more efficient.Footnote 45
However, after a false start in the 1990s,Footnote 46 only recently have tax law scholars begun conceptualising data as autonomously taxable wealth.Footnote 47 More generally, authors do not thematise these challenges within constitutional-theoretical frameworks. However, the relationship between digital technologies and constitutional law goes beyond the protection of the integrity of the ‘free marketplace of ideas’, the guarantee of ‘digital rights’, or more broadly, the political public sphere and market competition. Instead, it touches upon the very sources of integration/legitimation of constitutional states as they emerged following the secularisation of modern societies.
Digital constitutionalism, for its part, is by now a relatively settled strand of constitutional scholarship.Footnote 48 Recent contributions have highlighted how digital constitutionalism is intrinsically characterised by different perspectives and, in this sense, represents a discursive field whose elements are both complementary and contradictory.Footnote 49 However, if it aims at overcoming some limits of the liberal political theory underlying modern constitutional law and avoiding overlooking issues of societal power, a normative digital constitutionalism is called to frame excessive datafication and social justice as parts of a single project concerning the relationship between the digital revolution, the economic system, and constitutional law.
In both analytical and normative terms, social justice and value distribution are pieces of a broader puzzle within any digital constitutionalism aiming to stand as a counter-power to different kinds of authoritarianism. In this sense, digital constitutionalism needs to be an economic constitutionalism.Footnote 50 Looking at the trajectory of modern constitutional state, this means considering and revitalising the normative elements of economic and industrial democracy,Footnote 51 currently largely missing in digital constitutionalism discourses. Indeed, while traditionally overlooked or abandoned by constitutional theorists of the past century, such elements have historically played a significant role in the reality of 19th-century constitutional states.
Taking into consideration the different perspectives shaping its identity, digital constitutionalism is called to address at least four issues as part of a comprehensive, coherent research and policy agenda: (1) the negative impact of excessive datafication on contemporary societies (critical approach to datafication); (2) the legal conceptualisation of data for the extraction and distribution of its value (data as an economic factor); (3) the design of data taxes (tax design); (4) the interaction of data taxation with other legal regimes and issues of social justice at both national and global levels (economic digital constitutionalism).
4. From a critical approach to datafication to economic digital constitutionalism
A. Critical approach to datafication, regulatory goals, and data taxation
Any digital constitutionalism that takes the ‘datafication question’ seriously requires a critical approach and an awareness of the externalities deriving from datafication itself. Such an approachFootnote 52 is necessary to expand the regulatory goals of data governance and entails several consequences.
First, it forces new problems to appear or reframes existing ones.Footnote 53 ‘Big data’ and digital technologies are not only a means for received models of governance of populations and subjects. The amassing, analysis, and mobilisation of hybrid data repositories and real-time data flows – primarily driven by the profit-maximisation compulsion of (informational) capitalismFootnote 54 and the power-maximisation compulsion of political systemsFootnote 55 – open to new and potentially dangerous forms of governance, a new ‘digital political economy’.Footnote 56 The combination and mutual reinforcement of private/commercial and public/political surveillance, powered by digital and algorithmic technologies, end up affecting or manipulating individuals, groups, and social systems alike.Footnote 57
In Habermasian language, the excessive datafication of society narrows the ‘life-world’ spaces within societies.Footnote 58 At the same time, it allows for their colonisation/corruption by rationalities – economic, political, but also scientific, medical, mass-mediatic, etc. – whose inner expansive tendencies are less and less constrained.Footnote 59 Datafication is not a problem ‘only’ to the extent it reinforces the power- and profit-accumulation imperatives of political and economic actors. It re-frames, constrains, and potentially corrupts the social processes which allow for the protection and free development of individuals, collective groups, and functional systems within society.Footnote 60
Second, a critical approach highlights other – actual or potential – harms, eg, the energy consumption and environmental degradation linked to the data economy;Footnote 61 the exploitation and invisibilisation of old and new forms of labour;Footnote 62 and the socioeconomic costs deriving from data- and automation-driven business models.Footnote 63 Such issues are hardly taken into consideration in data governance strategies. The latter – typically focusing on harms to individuals and issues related to data protection, hate speech, discrimination, misinformation,Footnote 64 market competition,Footnote 65 and, more recently, (urban) safety and sustainability – do not include the reduction of the compulsion to datafication among their goals. Instead, they aim to increase the availability, quality, sharing of dataFootnote 66 or, at best, the participation of data subjects in its control/management.Footnote 67 Despite its potential,Footnote 68 the principle of data minimization – one of the cornerstones of data protection lawFootnote 69 – remains underenforcedFootnote 70 and, in any case, limited to the relatively narrow realm of personal data.Footnote 71 What was designed as one of its primary safeguards – the possibility to refuse or withdraw individual consent to data treatmentFootnote 72 – has long shown its limits, especially in the age of so-called digital resignation.Footnote 73 Moreover, AI systems and intensive data processing put into question the very dualism between personal data/non-personal data on which data protection law has been built.Footnote 74
More generally, by creating ‘data markets’ controlled by data intermediaries;Footnote 75 by circumventing or softening fundamental right guarantees concerning mass surveillance, often in the name of child protection;Footnote 76 by encouraging ‘smart’ urban planning;Footnote 77 and by relying on the mirage of data anonymisation,Footnote 78 more recent regulatory instruments incentivise data (over)production, commercial surveillance, and stabilisation of dominant market positions of ‘data producers’.Footnote 79
The challenges underlying data governance do not derive only from the role of business actors in highly concentrated markets, actors that can often escape ‘hard’ regulation and create their own regulatory standards.Footnote 80 They also come from the fact that regulatory approaches based on data as economic good are ‘hardwired to only produce governance strategies that will facilitate the provision of more or better-quality data’,Footnote 81 thus overlooking other societal goals beyond data provision. Even risk-based regulatory approachesFootnote 82 and ‘hard’ prohibitions of more recent instrumentsFootnote 83 – pursuing goals such as protection of the democratic process, innovation, privacy, or minors’ wellbeing – have only limited effect if the internal incentives to excessive datafication are unaffected.
The reduction of surveillance-based, profit-driven, attention-maximising datafication can be pursued without foreclosing the prospects for authentic inclusion, participation,Footnote 84 solidarity, and emancipation brought by digital innovation, as well as the economic opportunities brought by safe digital technology.Footnote 85 Certainly, datafication of contemporary societies, on the one hand, and the emergence of informational capitalism, on the other, have been deeply entangled. But such entanglement remains contingent: technological progress does not necessarily need to consolidate the institutions and actors of (informational) capitalism and may be directed towards authentic socio-political change, whatever the latter might be. A digital constitutionalism that (re-)asks the ‘social question’ is not about fighting rear-guard battles. Rather, it is about reducing the incentives to the excessive datafication of society and the closely related pressure to societal manipulation/colonisation, which, under current modes of production, is in turn linked to other perverse dynamics of global neoliberal capitalism (especially the free flow of capital and the financialisation of the economy).Footnote 86 Here, tax policies represent an overlooked instrument.Footnote 87
Historically, taxes – especially progressive onesFootnote 88 – have served multiple goals, well beyond the ‘mere’ funding of governments.Footnote 89 Taxes may be imposed to reduce risks associated with lawful but potentially harmful activities, to tackle the societal power coming from the accumulation of profit by economic actors, to enhance social mobility, inclusion and purchasing power of low- and middle-income population.
However, the normative justifications of proposals for ‘digital services taxes’ – in the EU or elsewhere – normally focus on the need to target new forms of intangible wealth that escape tax systems (so-called base erosion) but rarely mention disincentivising excessive datafication as such.Footnote 90 Once again, this is a regulatory blind spot, showing a persisting market-driven imprint.Footnote 91 As long as the contrast to excessive datafication does not become a goal as such, any regulatory strategy is doomed to overlook the related risks or, worse, contribute to their invisibility.
Against this background, tax law is only one instrument of a necessarily diverse regulatory mix that also needs strategically to include both state law – notably corporate, labour, administrative, tort lawFootnote 92 – and non-state normative systems (international law,Footnote 93 but also social and private norms of different kindFootnote 94). As long as it makes the data economy – which, as we will see below, is primarily a rentier economyFootnote 95 – less profitable, reduces the compulsion to excessive datafication, and slows down automation,Footnote 96 taxation is a useful and, so far, relatively underexplored tool to tackle the negative externalities of data capitalism.Footnote 97
To summarise, a critical approach helps identify datafication as a self-standing problem and orient the related regulatory strategies. Here, one needs to consider the potential contradiction between conflicting regulatory goals of data taxation: budget-funding/distribution of value, on the one hand; reducing/disincentivising excessive datafication, on the other. Indeed, any data tax effectively reducing the datafication compulsion might shrink the tax baseFootnote 98 of data companies and therefore the targeted revenue.Footnote 99 Certainly, which goal should prevail is a policy issue that needs to be addressed through politically legitimated decision-making processes. However, thematising the reduction of datafication as a regulatory goal is worth in itself, as it makes the associated dangers more visible – it forces them to appear as a problem, opens to more informed debates, and re-politicises data governance.Footnote 100 Further, the automatic link between data taxes effectively reducing datafication and the shrinking of the tax base is not to be taken for granted and is so far supported by little empirical evidence. There is indeed the possibility – and initial evidenceFootnote 101 – that raising the marginal costs of data collection might push companies to pass the burden to users by charging them for using digital services that are currently ‘free’. This dynamic could bring people to use less of it or, more realistically, re-expand the base of traditional income tax – the one most affected by the ‘base erosion’ linked to the data economyFootnote 102 – thus potentially increasing tax revenues.Footnote 103
Be as it may, before elaborating concrete tax designs, scholars and policymakers need to explore understandings of data providing normative justifications to data taxation also in light of constitutional limits related to taxpayers’ ability to pay. In other words, any effective data taxation policy needs to frame data as a factor contributing to data-collectors’ tax base.Footnote 104 In this way, a critical approach to datafication represents the point of departure towards topics investigated by LPE scholarship and institutional economics.Footnote 105
B. Data as a legitimate tax base within a rentier economy
Today, it is almost a cliché to observe that data is the new oilFootnote 106 or, as the EU Commission put it, ‘the lifeblood of economic development’.Footnote 107 Besides more or less questionable metaphors, data has undoubtedly become a major economic factor in contemporary economies. Profits of business actors – both dominant and non-dominant – heavily rely on their capacity to access, extract, process, and monetise data.Footnote 108 Some scholars even argue that data might progressively replace money or at least some of its functions.Footnote 109
From a legal point of view, data is not a ‘thing’, as it is normally defined as a ‘digital representation of information’.Footnote 110 Law ‘codes’ data as a ‘representation’, constantly replicated for a potentially infinite number of times. The term ‘data’ clusters a constellation of activities, notably inputting, processing, organisation, abstraction, units, aggregation, and resourcing of information through digital technologies.Footnote 111 But the law does not only define data. It also contributes to making it an economic factor.
Taking a step back, one can observe that, from an economic point of view, data becomes a valuable good through the aggregation and analysis of pre-existing information in huge amounts, via technologies created by software engineers and data scientists.Footnote 112 This constellation of activities, however, is difficult to evaluate economically, as it is contingent on specific technoscientific, legal, and institutional ecosystems.Footnote 113
Indeed, legal scholars have proposed numerous conceptualisations which, according to different policy goals,Footnote 114 variably frame data as object of property rights, privacy rights, or as commons.Footnote 115 However, it is crucial to stress the role of law in shaping the capacity of business actors to produce, commodify, and monetise data. The law ‘codes’ data as capital by making priority of use, durability, convertibility, and universality possible.Footnote 116 Law – or rather, a combination of (intellectual) property, contract, collateral, trust, corporate, bankruptcy, procurement lawFootnote 117 – contributes to this result in three ways.
First, the law keeps the ‘raw’ information potentially representable or processable via digital technologies a free, abundant resource, open to use, ‘mining’, even ‘scraping’.Footnote 118 Large networks of users ‘hooked’Footnote 119 to digital platforms and vast amounts of publicly available, ‘free’ information build the datasets necessary for the business model of data companies,Footnote 120 both those providing social media services and those focusing on AI.Footnote 121 Information potentially representable and/or processable via digital technologies is treated as res nullius rather than res communis.Footnote 122 But data is an economically relevant factor only above certain thresholds that can be reached only by (few) governmentsFootnote 123 and companies through their control of digital infrastructures.Footnote 124 In this sense, the data economy is an economy of scale.Footnote 125
Second, the law protects and stabilizes – notably via proprietary technologiesFootnote 126 – the unequal control over digital and computational infrastructuresFootnote 127 and, therefore, business actors’ capacity to appropriate and process information, excluding others from the use of data,Footnote 128 primarily through trade secrecy protections.Footnote 129 In theory, data could be an (impure) public, universally accessible good only with the combined presence of some elements: no or minimal constraints to computing and processing power; digital infrastructure built with interoperable standards; open-source (or unlicensed and in the public domain) software.Footnote 130 Absent such elements, data becomes de facto privatised.
Third, the law ensures the possibility of exchanging data for some direct or indirect advantage: direct monetisation, improvement/optimisation of one company’s own services, and influence on the social and political environment.Footnote 131
To sum up, data’s economic value varies depending on who uses it and how it is being used.Footnote 132 One feature – possibly the feature – of informational capitalism is that data is an economic good but its contribution to taxpayers’ wealth oscillates and is extremely difficult to pre-determine in monetary terms.Footnote 133 This makes it difficult to individuate its role in data-producers’ ability to payFootnote 134 and therefore a pre-determined tax base to target.Footnote 135 Also for this reason, data does not appear in corporate balance sheets and legal systems do not ‘code’ it as an economic good that can be taxed as such. This could be seen as an obstacle to taxing data in the light of the constitutional guarantees requiring taxes to be limited by taxpayers’ ability to pay and tied to some pre-determined tax base.Footnote 136
However, this obstacle also derives from the almost exclusive focus on the exchange value of data, that is, the possibility to convert it into other economic goods, notably money. Relatedly, in most advanced economies, the bulk of taxes presently falls on production, income, and consumption rather than rents. Therefore, tax policies are normally evaluated through the relatively narrow lens of the income tax as a default taxation model.
In contrast, recent scholarship has highlighted the use value of data, that is, the ‘value of being able to infer or predict likely future actions or effects’.Footnote 137 The use value of data, in other words, consists in exponentially enhancing data companies’ predictive power.Footnote 138 Predictive power and algorithmic control over user attention and digital infrastructures enable data companies to extract a rent Footnote 139 from the actors in their ecosystems (users, suppliers, and advertisers).Footnote 140 Even without an immediate monetary benefit – and often engaging in non-profitable or loss-taking investmentsFootnote 141 – acquiring predictive power provides structural economic advantages, further incentivising datafication.Footnote 142 The data economy, in other words, is primarily – although not exclusivelyFootnote 143 – a rentier economy.Footnote 144
The individuation of data and the predictive power it brings as an autonomously relevant economic factor – more specifically, as a source of rent – offers a normative justification for data taxation even in the light of constitutional norms imposing the pre-determination of a tax base as a limit to governments’ taxation power.Footnote 145
Further justifications may be figured out by looking at the process of data production. As seen above, this process involves the ‘mining’ of publicly available information and its subsequent transformation into effectively enclosed data. In this process, the intellectual labour and the technology made possible by data science are crucial.Footnote 146 Based on these observations and the scholarship supporting the existence of a right to science in international human rights law,Footnote 147 one could resort to the following argument: to the extent it dispossesses different types of workers of (the value of) their labour, without necessarily compensating such loss with a commensurate increase of aggregated demand, investment, and job gains;Footnote 148 and to the extent it is based on a common good – science – whose benefits are the object of an international right, profit-driven datafication provides data companies with a rent.Footnote 149
Put differently, data contributes to the tax base – and taxing data can be considered legitimate under traditional constitutional guarantees – not only because, above certain thresholds, it gives companies a crucial economic advantage through the use value/predictive power but also because it structurally deprives other actors – notably workers, households, economic actors in subaltern market position, and humankind more generally – of economic opportunities and benefits to which they are also legally entitled. Either way, the data economy emerges as a typical rentier economy, encouraging further rent-seeking or ‘conspicuous consumption’ rather than productive investment.Footnote 150
These considerations relate to the objection that, because of its elusive monetary/exchange value, data does not per se contribute to the data collector’s ability to pay and cannot be used to determine a tax base. To be sure, several issues are left unaddressed, although, as we will see below,Footnote 151 such considerations have an impact on tax design.
Here, one needs to stress that grounding data taxation on data’s use value is compatible with different (legal) notions of data. In other words, as far as taxation goes, qualifying data as an economic factor is compatible with different legal qualifications (personal and non-personal, sensitive and non-sensitive, ‘simple’ data and metadata, ‘synthetic’ and ‘authentic’, etc.) and regimes (ownership, commons, etc.) inspired by distinct regulatory goals (protection of privacy and individual self-determination, market competition, well-functioning political sphere, intellectual property, etc).
C. A constitutionalist data taxation: Pigouvian, progressive, rent-targeting
As seen above, the regulatory goals of a ‘constitutionalist’ data taxation (may) include the reduction of excessive datafication, as well as issues of social justice, distribution, and legitimation of constitutional states. The ultimate choices concerning tax design are political and should be open to debate, struggle, conflict, and collective decision-making. Digital constitutionalism does not and should not offer a pre-made, ready-to-go, depoliticised policy agenda, be it economic or not.Footnote 152 Its normative nature, if any, does not consist in providing a pre-set project of a ‘good society’ but rather in highlighting the crucial role of some aspects in policy fields where ultimate decisions will still be based on the interaction and conflict among politically engaged, strategic societal actors. In this sense, what a ‘good’ digital revolution is – or will be – can only result from this ‘blind’ collective dynamic. However, the arguments concerning the ability-to-pay objection are not neutral in terms of tax design.
Firstly, and more generally, any tax design coherent with the regulatory goals of digital constitutionalism needs to incorporate Pigouvian elements. This means that data taxes should target data companies and other actors of the data economy as actors who produce externalities and be designed to change their behaviour.Footnote 153
Secondly, data taxation informed by digital constitutionalism includes elements of progressivity.Footnote 154 In data capitalism, any form of data taxation has structurally progressive effects on the overall tax system, regardless of whether the specific tax imposition is designed as actually progressive or not, with increasingly higher ratesFootnote 155. While ‘data-rich’ taxpayers – basically, big data companies – easily avoid traditional income taxes,Footnote 156 low- and middle-income (‘data-poor’) taxpayers (eg, the individual Internet user) cannot. Therefore, via an exemption threshold calculated based on the volume of data traffic,Footnote 157 a fiscal imposition targeting (the collection and transmission of) data could easily capture only high-volume users that otherwise avoid income tax. Creating a ‘no-data tax’ area below specific data traffic volumes is not only sound in policy terms.Footnote 158 It is also a way to make data taxation compatible with relevant constitutional principles, given the specific features of the data economy. If the ability-to-pay principle can be respected only based on the use value/predictive power which comes from the amassing of huge amounts of data – typically those reached by data companies in an economy of scale – then a ‘no-data tax’ area for small-time data producers is a constitutional requisite for any legitimate form of data taxation.Footnote 159
Besides that, specific fiscal impositions could easily be made progressive. Once certain thresholds are crossed, the marginal tax rate could be designed as a logarithmic function to streamline data tax and prevent cliff effects.Footnote 160 The more data is produced, the (gradually) higher the tax rates and brackets would be. Once again, making data taxation progressive is not related only to the options concerning the distribution of the value extracted by oligopolistic economic actors. It also relates to underlying policy goals deriving from a digital constitutionalist agenda aware of the effects of excessive datafication as such and, at the same time, of the potential for inclusive economic growth coming from digital innovation.Footnote 161 Put otherwise, disincentivising excessive datafication means designing data taxes not as ‘flat taxes’ but rather as progressive ones.
Thirdly, a constitutionalist purview provides reasons to design data taxes as direct taxes targeting rent-providing positions.Footnote 162 For example, data taxes could be designed as royalties rather than digital services taxes (DSTs).Footnote 163 Royalties are fiscal impositions related to the extraction of (often) public resources. Those wishing to extract the resource must pay a ‘concession’ to the public authority for the right to do so. While not a tax in the technical sense, a royalty is commonly used by governments to raise revenue from extractive industries. Designing data taxes as royalties fits the reality of the data economy – as described above – taking into consideration the critical role of governments in developing digital technologies and infrastructureFootnote 164 as well as their continuing control over most computational and data infrastructures. Besides that, designing a data tax as a royalty has further advantages.
First, royalties do not require the activity to be productive and therefore are well suited to target sources of wealth that provide only potential rents.Footnote 165 As they do not target income deriving from data but data use as such, they bypass the problem of the delayed income realisation typical of the data economy.Footnote 166 Second, contrary to excises,Footnote 167 royalties can easily be made progressive – thus increasing based on some production measurement (eg, data traffic volumes) – and the related burden cannot easily be transferred to consumers. Third, contrary to DSTs, legal bases for adopting royalties can be found in (an evolutive reading of) existing international tax and trade treaties, which allow withholding taxes at source such as are applied to royalties. Therefore, they might forestall retaliatory measures, notably by the US, and reduce the need for international cooperation.Footnote 168
Even the criticism normally raised against royalties in the industries where they are most common – that they deter investment – does not necessarily apply to the data economy. Progressive royalties on data would only deter a specific business model. Moreover, the political economy surrounding data ensures that the ‘raw’ information potentially representable or processable via digital technologies, contrary to oil, remains a free, abundant resource, open to use and ‘mining’ even by non-profit actors. As highlighted above, the revenues of a few oligopolistic data companies are more akin to a rent and hardly create new value as such.
Fourthly and finally, a digital constitutionalist approach does not only aim at reducing excessive datafication and distributing wealth but also at countering the ‘governability’ and legitimacy problems of contemporary constitutional states accelerated by the digital revolution.Footnote 169 To that purpose, the related economic benefits should not necessarily be directed to individuals whose information was used to accumulate data (so-called data dividend taxes).Footnote 170 Indeed, data dividends feature some of the same issues as income taxation, that is, the problem of assessing the value of data collected from a specific territory, with the valuation and localisation problems coming with it. Further, the data economy is not based only on personal data but also – and, after the AI revolution, mainly – on non-personal, synthetic, and meta-data. Most importantly, data dividends de-responsibilise governments and de-politicise data governance.Footnote 171
Put otherwise, deciding how to specifically distribute the value produced in the data economy through the institutions of the political system potentially re-legitimises constitutional states and strengthens their role within (global) economic governance. Contrary to data dividends, direct taxes and royalties on data give governments the opportunity to regain legitimacy precisely because they do not take away from them the burden of adopting collectively binding decisions over different distributive options.
However, such decisions must consider that the data economy is intrinsically transnational. An authentically normative digital constitutionalism needs to thematise distributive issues through the lens of states’ legitimacy and within the framework of global (economic) governance.
D. Economic digital constitutionalism within global governance
In the globally dispersed data economy, location of income and ownership of data mean little to tax purposes. Relatedly, the externalities of datafication are also highly dispersed. Decisions concerning data governance, the contrast to excessive datafication, and distributive justice need to take these elements into account,Footnote 172 also from the perspective of the North/South divide. Such issues have already been debated for a long timeFootnote 173 and go well beyond the scope of this Article. This section highlights their analytical and normative interconnectedness from the perspective of digital constitutionalism, especially if the latter aims to hold onto its ‘global’ natureFootnote 174 and does not want to reinforce hegemonic positions within world society.Footnote 175
Taking a step back to tax design, one can observe that data taxes designed as excises, direct taxes, or royalties can be purely domestic. They do not target income and, therefore, bypass the problems of income realisation/localisation with which tax cooperation initiatives have been strugglingFootnote 176 for over a decade.Footnote 177 For the same reasons, such taxes are less likely to trigger trade conflicts.Footnote 178
This notwithstanding, global tax cooperation still needs to be thematised within a digital constitutionalist framework, especially in light of trends towards the liberalisation of data trade in international trade lawFootnote 179 and the persisting unequal control over digital infrastructures.Footnote 180 Overlooking such elements by designing purely domestic taxes targeting data traffic individuated through digital infrastructure may be convenient to Global North countries and the main competitor of the US for global hegemony – China.Footnote 181 However, data taxation policies overlooking issues of international cooperation and the North/South divide risk becoming a redistributive endeavour involving already-hegemonic actors that does not tackle and might actually reinforce digital colonialism.Footnote 182
At the same time, when it comes to distributive obligations with respect to taxing rights, it is crucial to differentiate between the collective and the individual level, as ‘the reallocation of taxing rights from state to state does not necessarily help when it comes to fulfilling duties of justice towards individuals’.Footnote 183 Put otherwise, the normative goals of an economic digital constitutionalism are not necessarily met by a value distribution from state to state but require a consideration of the actual political decisions over the distributive choices towards both individuals and communities.
Such issues are linked to the role of global institutions, notably international organisations and transnational networks.Footnote 184 This consideration opens to issues of institutional design and management of conflicts across different governance regimes.Footnote 185 Put differently, an effective constitutionalisation of the digital sphere can – has to – involve institutions that compensate for the absence of a world state (and tax authority)Footnote 186 by collectively fulfilling the functions performed by political systems at the global/transnational level.Footnote 187
As recalled, institutions such as the OECD have become incubators of global tax cooperation – especially in the field of digital economy – and should be considered at least as important as those traditionally at the forefront of digital constitutionalism’s reflections (eg, the ICANN). Whether such efforts will be successful/effective or not is an open question. At the moment, the multilateral anti-BEPS initiativeFootnote 188 seems to have stalled, potentially giving rise to new trade conflicts.Footnote 189 However, such fora allow for the thematisation and, potentially, contestation of key issues of economic global governance.Footnote 190 From this perspective, the recent emergence of a competing initiative within global tax governanceFootnote 191 might be seen as a positive development not only from an experimentalist perspective but also because it contributes to the further politicisation of such regime. The purposes of re-legitimation of political institutions and global justice impose a political reflection on how to direct the related revenue.
Indeed, besides the more general goal of tackling the fiscal crisis of states,Footnote 192 it is crucial to strengthen the social fields – eg, medicine, information/press, science – most affected by the commodification trends and the reduction of aggregated demand deriving from the business models currently dominating the data economy.Footnote 193 If digital constitutionalism wants to live up to the aspirations of any authentic constitutionalism and constrain the expansive tendencies of economic systems ‘freed’ by the digital revolution, it has to embed forms of social protection,Footnote 194 especially for the processes instrumental to political participation,Footnote 195 socioeconomic inclusion, and generation of socially shared consensus, knowledge, truth. Moreover, focusing on the protection of social processes as such might help address two admittedly difficult regulatory questions: first, the choice between the collective and the individual levels of revenue reallocation/distribution; second, the individuation of the point where datafication becomes ‘excessive’.Footnote 196
In this regard, scholars have formulated proposals concerning, for example, funding some form of universal basic income (UBI).Footnote 197 Other contributions explore the possibility of using the data tax revenue to fund open medical research, supranational organisations such as the EU, especially considering their role in tackling other pressing global governance issues such as the climate emergency.Footnote 198 But many other solutions and proposals might be advanced, based for example on the environmental, social, and governance (ESG) agenda promoted by the United Nations Environment Programme Finance Initiative;Footnote 199 or – even more interesting and disruptive – the Program of Action on the Construction of a New International Economic Order elaborated by the Progressive International,Footnote 200 published on the occasion of the 50th anniversary of the UN General Assembly resolutions adopted in 1974 that aimed at establishing a New International Economic Order (NIEO).
Before concluding this section, one needs to address another question: Why ‘economic digital constitutionalism’ instead of the less conceptually demanding ‘digital constitutionalism for the economy’ or ‘constitutionalism for the digital economy’? And why is taxation important?
By using that phrase, the purpose is to link the analytical and normative frameworks of digital constitutionalism to the older conceptual tradition of the economic constitution, which, in its different understandings, goes beyond the ‘mere’ constitutional governance or regulation of the economy.Footnote 201 Importantly, that phrase brings back an entire set of normative elements related to forms of democratic legitimacy and decision-making that, instead of exclusively focusing on the traditional schemes of representative democracy turning around formal political institutions and procedures, take the condition of subaltern groups – especially within economic relationships – as the starting point to build any effective form of constitutional legitimacy and normativity even in the digital age.Footnote 202
Relatedly, arguing that data taxation is a crucial instrument of a properly normative and therefore economic digital constitutionalism is meant to highlight two elements. First, data taxation aims – should aim – at constraining the expansive/colonising tendencies of both economic and political systems ‘freed’ by datafication processes. Second, value distribution through taxation may ultimately (re)legitimise the institutions of the political system, ‘exhausted’ by the persisting necessity to produce collectively binding decisions even when social consensus declines or is absent.
In other words, and without necessarily subscribing to the famous dictum of the US Supreme Court ‘the power to tax involves the power to destroy’,Footnote 203 data taxation could be a crucial element to re-assert the centrality of political institutions in global governance, especially in the light of the historically genetic link between taxation, democracy, (popular) sovereignty, and modern constitutionalism.Footnote 204
Along the same lines, advancing tax-based policies and proposals does not necessarily mean validating or naturalising governance approaches treating systemic problems in global capitalism as externalities and the results of failures of otherwise ‘perfect’ systems. The proposals mentioned above are rough and need to be discussed, analysed, and tested, also based on empirical elements provided by other disciplines. In no way are they supposed to be exhaustive, and, most importantly, they do not exclude regulatory measures of different kinds. In particular, they do not rule more radical options out, for example those questioning the current governance of digital infrastructures, one of the key structures underlying the data economy itself.Footnote 205
5. Conclusion: a call for normative and institutional imagination
An authentically normative digital constitutionalism may point at some common principlesFootnote 206 and regulatory directions but, as such, is not and should not be a post-political discourse offering a pre-made agenda.Footnote 207 Neither constitutional law (theory) nor law in general have the communicative/symbolic potential to solve, by themselves, the material challenges of the digital revolution and global governance. However, recognising the role of the law in constituting the institutions of (digital) capitalism – like the recent LPE scholarship does – is a fundamental step to address the related challenges and contribute to their transformation.Footnote 208 At the same time, the transformative possibilities of the law should not be overestimated. This holds particularly true when it comes to more radical forms of contestation of global economic (re)production.
Digital constitutionalism can and must provide a discursive field to thematise constitutionally significant economic issues: a ground for conflict and collisions to arise and produce constituting and limiting norms.Footnote 209 Opening the gates to conflict – forcing the problems to appear as problems – strengthens the analytical and normative potential of digital constitutionalism. It also establishes fruitful conversations with other political–economic discourses and practices concerning the material conditions of the digital revolution at the global level, which, in turn, might be enriched by more direct interactions with explicitly constitutionalist frameworks.
To be sure, the cluster of informational capitalism,Footnote 210 ‘technofeudalism’,Footnote 211 data colonialism,Footnote 212 and different forms of both public and private digital authoritarianismFootnote 213 will hardly be overcome only through litigation before (constitutional) courts. Nevertheless, collectively rethinking the meaning of ‘(un)constitutional’ in the digital era is a necessary step towards radical transformation. Digital constitutionalism can emerge as a counter-discourse and be authentically normative if and to the extent it (1) breaks the conflation between a ‘public sector’ driven by digitalised power and a ‘private sector’ driven by digitalised profit;Footnote 214 and (2) contributes to the re-legitimisation of both national and international political institutions, reducing their tendency to decomplexify their social environments through techno-authoritarian solutions.Footnote 215
In both cases, data taxation should be considered as a piece of a complex puzzle whose contours (should) remain open to institutional and legal imagination.Footnote 216
Supplementary material
The supplementary material for this Article can be found at https://doi.org/10.1017/elo.2024.37
Competing interests
The author has no conflicts of interest to declare.
Funding and acknowledgements
This Article is part of the research project ‘The principle of progressivity in the digital economy. Possibility for a “Data Capital Tax” in Italian and European constitutional law’ funded by the Italian Ministry of University and Research within the ‘Programma per Giovani Ricercatori “Rita Levi Montalcini - Bando 2020 - project code MUR: PGR20TV1NP‘. The research was entirely conducted while I was Assistant Professor at University of Trento School of Law in 2023-2024. For comments, suggestions, and criticisms, I am indebted to Anna Beckers, Marco Dani, Giovanni De Gregorio, Ernst Ekkehard, Melanie Fink, Chiara Gentile, Rachel Griffin, Ylenia Guerra, Raffaela Kunz, Julieta Lobato, Agustín J. Menéndez Menéndez, Jacopo Ratti, Mona Sloane, Thomas Streinz. I would also like to thank the participants in the workshop ‘Rethinking Science in Times of Open Science?’ (Collegium Helveticum – ETH Zürich, 21 March 2023); in the panel ‘Power Circuits: Intersections of Digital Infrastructures and Constitutions’ of the Research Network Young Digital Law Conference 2023 (Department of Innovation and Digitalisation in Law, University of Vienna, 5 July 2023); in the online workshop ‘Taxing Data as an Instrument of Economic Digital Constitutionalism’ (15 November 2023); and in the Law & Tech seminar ‘Taxing Data as an Instrument of Economic Digital Constitutionalism’ (Católica Global School of Law, Lisbon, 16 May 2024). All errors remain my own. All websites and online sources have last been accessed on 1 February 2025 if not otherwise indicated.
Open access
