[REPLAY] Standford winter webinar with Florence G’sell on “The Impact of Blockchains for Human Rights, Democracy and the Rule of Law”
18 January 2023
[REPLAY] Content Moderation in the Age of DSA
13 February 2023
By Tamian Derivry
One of the key objectives of the EU’s 2020 recovery plan is to achieve the green and digital “twin” transitions. Not only do both these transitions happen to be priorities for the EU, they also overlap in an increasing number of ways in what European Commissioner for Internal Market Thierry Breton (2021) called the “green-digital nexus”. So far, the EU has largely focused on how digital technologies can be used to accelerate the green transition, e.g., by improving the efficiency of energy-intensive industries, through initiatives like the European Green Digital Coalition. Yet growing concerns have been raised about digital pollution, or the negative effects of the digital sector on the environment.
In recent years, the EU’s digital policy has centred on the notion of digital sovereignty, which encompasses a wide variety of policies, including the much-discussed Digital Services Act (DSA) and Digital Markets Act (DMA). At EU-level, the term was greatly influenced by the French concept of strategic autonomy, or the ability to control critical sectors of the economy and ensure national security. The role of the state in enabling individual sovereignty within the digital space, or the self-determination of individuals over their online experiences, has also been stressed by Germany. In fact, the EU policy discourse has often framed technological independence as instrumental to the protection of European values and fundamental rights, somehow reconciling the two approaches.
Digital sovereignty and environmental protection have been categorised as distinct agendas in the EU, despite criticisms by both private actors like the Digital SME Alliance, and academic actors such as the German research project CO:DINA. The term sustainable digitalisation is increasingly being used to highlight the need for digital policies to address not only the social impact of digital technologies but also their impact on the environment. This article focuses on the intersections between digital sovereignty and digital pollution. First, outlining some of the core aspects of digital pollution and the policy response in the EU, it will go on to address the intersections between digital sovereignty and green technology and conclude with some remarks on possible tensions with “digital sobriety”.
I. The digital sector, digitalisation and environmental protection
The focus of political debate has largely been on the virtual spaces that digital technologies create and the socio-political challenges that arise from them: disinformation, surveillance, hate speech, etc. As Guillaume Pitron points out in his investigation into the “journey of a like”, it is essential to realise that these virtual spaces are built on very material infrastructures – routers, submarine cables, data centres, etc. – and require often scarce natural resources – cobalt, lithium etc. – in order to fully understand their environmental impact. But what is this impact and what is being done to address it?
(1). Measuring and understanding digital pollution
Methodologies differ greatly in measuring the footprint of the digital sector and, more importantly, there is no clear-cut definition of what the “digital sector” is. Conventionally, it has been the sum of data centres, transmission networks and user equipment. But increasingly, there have been overlap with other sectors (e.g. in the automotive sector, navigation assistants) and there is a growing number of gadgets that are not always included, which causes the sector’s footprint to be either underestimated or overestimated.
Thus, it is useful to look at the results of a range of studies. A meta-study by researchers at Lancaster University, which corrected for errors in existing estimates, found that the digital sector’s carbon footprint accounts for 2.1 to 3.9% of global emissions in 2020. In France, a 2022 report by the French Environment and Energy Management Agency (ADEME) and the Telecommunications Regulatory Authority (ARCEP) found that digital technology currently accounts for 10% of France’s electricity consumption and 2.5% of its carbon footprint. In all these studies, there is uncertainty due to the difficulty of obtaining data from manufacturers.
Drawing on the results of older studies, the Lancaster researchers argue that ICT’s carbon footprint has increased by at least 40% between 2002–2012, or roughly twice as fast as global emissions. In fact, the digital sector has one of the fastest growing footprints in the economy. According to a 2019 report by the Shift Project, a think tank advocating for “digital sobriety” and low-tech, the sector’s greenhouse gas emissions will significantly increase in the future: from 3.7% in 2019 to 8% in 2025. However, these predictions are highly uncertain due to the unpredictability of certain key variables that can have negative or positive effects.
Despite these worrying figures, the idea that digital technologies provide net benefits for the environment has remained strong in business and policy circles. This could very well be the case as a result of their so-called indirect effects, e.g., digital technologies enabling remote work and therefore diminishing mobility pollution. Yet the significance of these effects should not be overstated, e.g. some jobs cannot be done remotely. Besides, there are also negative indirect effects, e.g., remote work might create new mobilities as in the case of digital nomads. Ultimately, there is general agreement among academics in the field that, regardless of the potential positive impact of digital technologies on the environment, the sector must actively reduce its negative effects.
The ADEME/ARCEP report finds that devices account for most of the impact (65-90%), followed by data centres (4-20%) and networks (4-13%) and that most of the impact occurs at the production phase, followed by the use phase, while the distribution and end-of-life phases have a minimal impact. Although there is some country variability, not least because of differences in the energy mix, this is consistent with international studies, such as that by Gupta et al. Additionally, they illustrate that carbon emissions from energy consumption are decreasing thanks to efficiency gains, while the overall carbon footprint of IT systems continues to grow. This is primarily due to the increase in hardware manufacturing and infrastructure, which is in turn driven by increased data and equipment use.
(2). Policy response to digital pollution in the EU
The focus of the EU has long been on the management of e-waste, which is one of the rare statistics on digital pollution provided at EU-level (according to Eurostat, only 46% of e-waste was collected for recycling in 2020). The EU has legislation in place for the management of e-waste, notably the Waste Electrical and Electronic Equipment (WEEE) Directive 2012 and the Restriction of the use of certain hazardous substances in electrical and electronic equipment (RoHS) Directive 2011, which are both currently under review.
Beyond e-waste and reuse, the significant impact of the production phase, especially of user devices, has drawn attention to so-called “eco-design” approaches, which aim to improve the sustainability of products throughout their life cycle, and not only at the end-life phase. In March 2022, the European Commission published a proposal for a Regulation on Ecodesign for Sustainable Products, which would extend the scope of the Ecodesign Directive to improve product durability, reliability, reusability, upgradability and reparability.
Further legislation is planned specifically for the digital sector under the “Circular Electronic Initiative”, which is part of the EU’s 2020 Circular Economy Action Plan. This includes the common charger proposal (as a revised version of the Radio Equipment Directive), on which an agreement was recently found between the EU Parliament and Council, and an “EU-wide take back scheme to return or sell back” devices. Other initiatives particularly relevant for the digital sector include the “right to repair”, for which an impact assessment is underway and which could possibly be implemented as a revision of the Sale of Goods Directive, and the proposed new battery regulation, on which an agreement is expected soon.
But as Gauthier Roussilhe argues, eco-design has limitations as it only tackles the supply side of digital pollution and can easily be used to justify environmentally unfriendly practices. New initiatives seek to address the demand side by improving consumer information. In March 2022, the European Commission proposed a directive on empowering consumers for the green transition (revising the Unfair Commercial Practices Directive and the Consumer Rights Directive). This directive would impose new obligations on businesses to provide consumers with information on the durability and reparability of products, which could help to protect them from greenwashing and premature obsolescence. Yet consumer information may not be enough to deal with the rapid growth in the use and purchase of digital equipment. At national level, awareness campaigns have been launched to nudge consumers towards retaining their devices longer and limiting their data consumption, for example in France, a joint campaign by the ADEME and the French government.
Nonetheless, these campaigns do not address emerging challenges, such as the exponential growth of energy consumption on the supply side of data storage, namely cloud providers and data centres. Although they do not account for a very significant share of digital pollution today, this share is likely to increase if nothing is done. The EU has already adopted an Ecodesign Regulation on servers and data storage products and is exploring other solutions. Discussions are also underway regarding the high energy consumption linked to cryptocurrency mining.
II. Complementarity between digital sovereignty and green technology
All these initiatives indicate that the EU is now actively seeking to address the issue of digital pollution, primarily through regulatory measures. These have been overwhelmingly framed as distinct from the digital sovereignty agenda. There are, however, untapped links between the two, which may be used to improve the effectiveness of green tech policies. Drawing on the work of Julia Pohle and Thorsten Thiel, three main areas of the EU’s current digital sovereignty agenda can be identified: industrial policy to ensure technological independence; competition policy to ensure a level playing field with foreign companies; data and platform governance to protect fundamental rights and values. This is a tentative list of the positive links between these three areas and the mitigation of digital pollution:
(1). Industrial policy: relocation and enforcement of environmental standards
European countries have been pointed out for their technological dependence to big US and Chinese technology companies throughout the supply chain. This stressed the need to shift from an almost exclusively regulatory approach to an ambitious digital industrial policy, for example by investing in the manufacturing of electronic equipment in the EU. In fact, some countries such as France have already announced important investment plans in research and the construction of microchip factories, even if European countries already risk being overtaken by the United States in this so-called “chip race”. In addition, calls by the private sector have triggered a discussion on the role of public procurement in supporting the European digital sector, including on the legal and economic implications of adopting a Buy European Tech Act.
The process of relocating and supporting European industry is perhaps the most straightforward link between digital sovereignty and the reduction of digital pollution. First, it could reduce the impact of product transportation, which, for example, accounts for 8% of Apple’s carbon footprint in 2021 according to the company’s environmental progress report. However, existing studies indicate that the distribution phase may not constitute a significant part of the environmental footprint of the sector as a whole. Second and more importantly, the relocation and Europeanisation of the digital sector is the most effective process whereby European governments can make sure that they are able to enforce their policies. Enforcement has been particularly challenging in the area of digital policy due to the cross-jurisdictional nature of large technology companies and the differences that exist between European and US digital law in particular. The improvement of the EU’s enforcement capacity similarly applies to environmental standards, particularly with regards to eco-design practices.
(2). Competition policy: market-driven adoption of environmental standards
The ability of European jurisdictions to enforce their standards is not the only factor fostering the adoption of environmental standards by technology companies. Another important factor is the state of the market. At present, it is dominated by a few large companies. For example, Samsung, Apple and Xiaomi alone are said to account for 78% of smartphone sales in Europe in 2021. Oligopoly situations are conducive to business practices harmful to the environment, such as in particular planned obsolescence, that is when products are designed to have shorter lives than they could, forcing customers to buy new items. In a competitive market, consumers can easily turn to more sustainable alternatives, and therefore companies have no interest in planned obsolescence.
There are already alternatives on the European market that claim to have developed more ethical and environmentally friendly products, such as the Dutch company Fairphone. In fact, some economists argue that planned obsolescence is possible in a competitive market because of information asymmetry between consumers and manufacturers. Attempts to improve consumer information may help to make the sector more transparent, thereby reducing this asymmetry.
However, beyond information asymmetry, the digital sector is particularly affected by lock-in effects, whereby dominant companies create special features that increase the cost of switching to a competing technology. This is the case for example of Apple chargers, which are designed to only function with Apple devices, thus increasing the cost of buying a non-Apple device. This in turn has a negative impact on the reusability of products, as customers are forced to buy a new charger when buying a new phone. Regulations aimed at limiting the anti-competitive practices of technology companies, which also aim to rebalance the market power of European companies, therefore have the potential to reduce unsustainable business practices. A good example is the European Commission’s proposal on the common charger, which promotes both a level-playing field with foreign technology companies and the reduction of e-waste.
(3). Data and platform governance: data minimisation and sustainable nudging
This last link is probably the least obvious and remains largely unexplored. The German research project CO:DINA recently published a study on the theoretical relationship between sustainability and data sovereignty, as the ability to decide on the purpose of data-driven applications. More specifically, they argue that increased control over the use of online data by state and civil society actors in the EU could help achieve data-driven environmental action. In addition, there might be a link between data protection and the reduction of data pollution. Article 5 of the GDPR (General Data Protection Regulation) provides that personal data collection is subject to the principles of data minimisation and storage limitation, i.e. is “limited to what is necessary in relation to the purposes for which [data] are processed”. However, there is little empirical evidence showing to what extent data sovereignty and data protection can be used to reduce data pollution.
CO:DINA published another study investigating the link between environmental protection and platform governance. Large digital platforms have acquired considerable power over our behaviour and consumption patterns through the algorithmic curation of information. They are often promoting online overconsumption and the spread of climate misinformation, which in this case, constitutes a negative indirect effect of digital technologies. The role of platforms in shaping (un)sustainable markets has been largely overlooked in the EU’s digital policy agenda. New obligations could be explored to provide consumers with environmentally relevant information about online services and products building on existing initiatives such as the Digital Product Passport, which aims to improve product traceability. In line with current attempts to regulate so-called “dark patterns” and disinformation, new obligations or incentives may also be explored with regards to the online filtering of products and services in order to limit the algorithmic promotion of unsustainable consumption or the dissemination of climate misinformation.
III. Is digital sovereignty and digital sobriety compatible?
The concept of digital sobriety was first defined by Frédéric Bordage, founder of GreenIT.fr, as “an approach that consists of limiting our digital usage and designing more sustainable digital services” (own translation). It was coined in French [sobriété numérique] and, as a quick search on Google Trends shows, still mainly resonates in France. In contrast to the notions of “sustainable” or “clean” tech, digital sobriety is not only about reducing the environmental footprint of digital products and services throughout their life cycle, but it also incorporates a behavioural dimension that seeks to limit our use of these products and services. It is advocated by several civil society organisations, notably the Shift Project.
Digital sobriety can conflict with the digital sovereignty agenda, but arguably not inherently. The main line of tension revolves around high-tech consumption, production and innovation, which can be seen as both instrumental in building independent technological capabilities and detrimental to the environment. Although high tech is becoming a strategic asset in international politics, it tends to be highly energy-intensive and requires rare materials, the extraction of which can be very polluting.
This tension is rooted in the divide between high-tech and low-tech approaches. The example of new modes of mobility can help illustrate this divide. Proponents of green or sustainable high-tech may say that autonomous cars will improve car efficiency, thus yielding net environmental benefits. Proponents of low-tech on the other hand, may respond that bicycles and trains still provide far greater net benefits than autonomous cars, and should therefore be the focus of modal change. In practice, this divide is often overstated, e.g., because not all forms of mobility can be replaced by trains or bicycles and because autonomous cars might be used to increase car-sharing rather than solo-driving, thereby reducing overall car production. Perhaps the most crucial contribution of the low-tech movement is to warn us against technological solutionism, the belief that technology will solve all our problems, and encourage us to question solutionist narratives, e.g. to look critically at the automotive industry’s promises of the environmental benefits afforded by autonomous cars.
In fact, it can be argued that uncovering technological solutionism, especially the great promises of the Silicon Valley, can help us reduce our technological dependence on foreign technologies and focus on supporting a European approach to sustainable technology, which reconciles low and high-tech. So far, the issue of digital pollution has been addressed mainly within the EU’s green agenda. But as the links I have highlighted suggest, it could greatly benefit from being further integrated into the digital sovereignty agenda. Ultimately, linking green technology and digital sovereignty policies more systematically constitutes another step towards making technology work for us and our environment.
Tamian Derivry is a double master’s student in European Affairs and Political Science between Sciences Po and the Freie Universität Berlin. He is also a research assistant at the Digital, Governance and Sovereignty Chair of Sciences Po.