The paper below is an academic and policy contribution supporting the European Commission’s NGI Commons project and current efforts to establish a “Digital Commons European Digital Infrastructure Consortium.” It is currently in a draft version available for expert feedback until 23 of July. A final version, edited and proofread, will be published in August 2024.

The paper aims to provide an overview of current policy debates that identify Digital Commons as potential providers of Public Digital Infrastructure and to highlight some key policy issues for the support of Digital Commons by public institutions by reviewing the literature on their interaction with the state. To this end, the below paper includes five case studies selected and analyzed by external contributors.

If you find it difficult to leave comments via PubPub, you can email your comments to one of the authors, jan(at)openfuture.eu.

Introduction

Most of the early information and communications infrastructures were developed in the 20th century, a period that saw the emergence of the "modern infrastructure ideal," characterized by the provision of infrastructure directly by governments or through publicly regulated monopolies. Strong government involvement also characterized the early history of digital infrastructure. Economist Mariana Mazzucato calls this involvement the "entrepreneurial state”1. Public sector contributions to foundational technologies and innovations such as microprocessors, the HTTP protocol, and GPS, as well as policies aimed at building the "information superhighways" on which our knowledge economies would be built, show how critical public involvement has been in the history of digital technologies.

In the late 1980s and 1990s, however, an important policy shift occurred that led to the increasing deregulation and privatization of many infrastructures. This movement was based on the neoliberal belief that public intervention should be limited to market failures - where private actors cannot make sufficient profit to provide goods and services. This neoliberal turn also affected digital infrastructures, initially the telecommunications sector responsible for providing access to the Internet, but more importantly the development of digital services on top of these infrastructures in the beginning of the 21st century. The result has been a digital landscape driven primarily by market forces, fostering rapid wealth accumulation and the emergence of dominant digital platforms that exploit network effects, user lock-in strategies and data extraction to extend their reach across all sectors of economies and societies. Platforms have gained extraordinary power and wealth by coordinating diverse and independent actors while extracting value from their interconnections2. This capability has enabled them to challenge traditional infrastructure. Their vertical integration strategy, coupled with recent technological advancements—particularly the development of centralized cloud computing architectures exemplified by hyperscalers and the rise of artificial intelligence (AI)—has further increased their infrastructural power. Companies like AWS, Microsoft Azure, and Google dominate cloud computing, creating bottlenecks in market segments and posing future risks of similar constraints in AI systems and applications3.

Today, global policies on digital technologies seem to be at a pivotal juncture. After years of a utilitarian, laissez-faire approach, and a reliance on market forces to dictate technological development, there is a growing recognition of the need to shape digitalization in alignment with societal priorities. As highlighted in the United Nations Secretary-General’s Roadmap for Digital Cooperation: “Digital technology does not exist in a vacuum – it has enormous potential for positive change, but can also reinforce and magnify existing fault lines and worsen economic and other inequalities. (...) The world is at a critical inflection point for technology governance.”4

At the same time, research on infrastructure shows that the traditional economic perspective, which limits public intervention to market failures, is increasingly being challenged. Taking into account the high social demand for public infrastructure and services, many voices advocate for a more generative role for public policy, focused on ensuring the quality and quantity of basic services, rather than focusing on areas where there is limited private supply of public goods5. This shift prompts a reevaluation of the classical economic divide between private and public spheres and a redefinition of the “publicness” of infrastructures. Indeed, arguments for the publicness of infrastructure are increasingly being examined critically by both academics and activists, asking what benefits infrastructure provides, who benefits from this infrastructure, what kinds of knowledge - expert or non-expert - define these benefits, and what technical tools are needed to achieve them6.

Over the past decade, there has been a significant adaptation in policies to address the new power dynamics created by dominant digital platforms. This shift recognizes the collective responsibility in steering technological trajectories and emphasizes the necessity of public interventions—not only through regulation but also through strategic investments aimed at reclaiming digital sovereignty. The European Union’s evolving role in the digital sphere reflects this broader global resurgence of industrial policy and the reconceptualization of the state as a strategic actor7. Regulations such as the Digital Markets Act (DMA) and the Digital Services Act (DSA) have formed an initial answer in this respect. More proactive approaches in shaping technological development are illustrated by recent initiatives such as the “Next Generation Internet”, supporting “open-source decentralized digital solutions” aimed at supporting values such as “openness, inclusivity, transparency, privacy, cooperation, and protection of data”8. While governments have historically invested heavily in the support of innovation and startup programs, or in the digitalization of small and medium Enterprises (SMEs) and industries, increasingly, political focus is turning towards alternative forms for the governance of technology, moving beyond market-driven models. This shift is far from being specific to the EU and has given rise globally a renewed interest in public digital infrastructure.

Infrastructure policies are usually the results of a specific and political problematization of social relations, that defines a hierarchy between elements that are perceived as foundational because they provide a generative input into a wide range of activities. This makes infrastructures recognized for their society-wide economic and social functions, even though their overall impact and spillover effects remain challenging to measure9. In this context, various political, cultural and legal visions regarding public digital infrastructure are co-existing today, forming a diversity of alternative models to the private dominance of today's digital landscape. The Government of India has been a strong advocate of a specific vision of what “Digital Public Infrastructure” (DPI) should be, inspired by its own experience in developing a stack of open and interoperable technologies that facilitate essential functions for both public and private service delivery. In this context, public institutions control protocols and norms to ensure interoperability while public-private partnerships are established to deploy public services on top of these building blocks10. Other visions for public digital infrastructure emphasize on the public responsibility in developing and maintaining an ecosystem of open protocols and software components that rely on collective governance and non-extractive economic models11.

Such visions for public digital infrastructure follow trends in the management of infrastructure challenging the traditional private/public dichotomy, and calls for the integration of various forms of collectivity into public infrastructure. Within this framework, public infrastructures should not only have public attributes and functions but also rely on public ownership, which can range from ensuring public control to public funding or even direct public provision. Commons - and particularly digital commons, are gaining prominence in this context, as they promote public participation through the involvement of citizens, civil society, and private entities in resource management12. The concept of digital commons has been used to describe the large diversity of digital systems and solutions that have been owned, developed and maintained by a group rather than by a single individual or entity. These groups differ from businesses or public institutions because they do not coordinate through pricing or subordination, but through collaboration among peers13. Digital commons often began as small community projects but many have scaled to become society-wide infrastructures, attracting vast contributor bases and enabling global applications and uses. Early 21st-century examples like Wikipedia, the online encyclopedia, and Apache, an open-source software that supports over 20% of the million busiest websites globally14, exemplify the profound impact of digital commons. Some authors have even described the current Internet stack to be an “accidental megastructure" of digital commons15.

In this context, digital commons are gaining political attention for their potential to provide an answer to the high social demand for digital infrastructure, and especially the demand for public digital infrastructures that would form alternatives to current private platforms and in that sense further enhance sovereignty and the respect for digital rights, but also create a more competition-friendly digital environment, that would improve the accessibility of services. In this context digital commons are perceived as “a key mechanism for providing public digital infrastructure”16. The report “Towards a Sovereign Digital Infrastructure of Commons” adopted by 19 EU Member States during the Digital Assembly held in July 2022, does not only recognize the potential of digital commons to “support European digital sovereignty”, but also emphasizes on their potential as an alternative to the “enclosure strategies” of both “governments and major digital services providers”. Such an approach, based on “collective intelligence and networking” or “multilateral governance” is presented as a form of “public-civic-private cooperation”17. These mechanisms could be considered a potential answer to contemporary infrastructure critics, who reject centralized and technocratic public-private partnerships as a way to provide public infrastructures and present the commons as a way to reclaim the governance over technical systems18. Such criticisms have been particularly vivid in the context of digital infrastructures, where government-led initiatives can create risks for the respect of human rights19.

In addition, collaboration with digital commons is also an opportunity for public institutions that often lack internal capacities, funding and expertise to provide better digital infrastructure and services. This corresponds to the literature on infrastructure studies that argue that commons emerge specifically in the context of an “infrastructural gap” or in “times of trouble”, where public institutions are absent. This historical dichotomy between the commons and the state is present in the traditional literature on commons, which still largely considers their interaction to be antagonistic. It also explains the limitations of digital commons as providers of public digital infrastructure. The fact that commons, as a resource management principle, are considered to be a more participatory alternative to the private and technocratic nature of “Big Infra” projects20, means that they also appear to be better suited for infrastructures that mostly rely on ongoing human contributions and maintenance efforts, rather than on heavy capital investment. Recent research contributions on digital commons in particular, however, showed that depending on the political context, the state can play a constructive role in protecting and even supporting digital commons. Evidence from recent experimentations in public service provision show how co-production between self-governed citizen communities with hierarchical public administration can lead to the establishment of hybrid forms of institutions21.

Objectives of this publication

The purpose of this paper is twofold. First, it aims to provide an overview of the history of policies that have made digital commons potential providers of public digital infrastructure. To this end, the first section will critically examine the concept of public digital infrastructure by reviewing recent economic and sociological infrastructure studies. This section will already show the extent to which commons have been presented as an alternative to capital-intensive and technocratically dominated "Big Infra" projects. The next section explores contemporary visions of the "publicness" of infrastructure, showing that there is a shift from state ownership to various forms of collective involvement in infrastructure, with a focus on maximizing the public value of infrastructure. Finally, these findings will be linked to current debates about the potential fields of public digital infrastructure. The mapping of these debates will be limited to the strictly digital spheres of public infrastructure, where digital commons are already providing alternative infrastructures or are part of public strategies to build and maintain them.

Second, this paper aims to identify key policy issues that need to be considered for the provision of public digital infrastructure through digital commons. In this context, it will review existing literature on the interplay between digital commons and the state to identify emerging issues and challenges in supporting digital commons as providers of public digital infrastructure. Through case studies of novel forms of public support for infrastructure provision through digital commons, the paper seeks to highlight approaches that address infrastructure gaps and contribute to a deeper understanding of how digital commons can sustain and enhance public digital infrastructure.

This paper is written as an academic contribution to improve our understanding of the potential public role of digital commons. It is also part of various contemporary efforts to include support for digital commons in public policy, especially in the European Union. It should contribute to the current objective of the European Commission to define a new strategic agenda for digital commons through the NGI Commons project22. It should also contribute to the design of the "Digital Commons European Digital Infrastructure Consortium"23, which is currently being scoped as an initiative of France, the Netherlands, Germany and Estonia.

What is Public Digital Infrastructure?

The contested notion of infrastructure

The word infrastructure was first used by French engineers working on public roads in the nineteenth century. It was used to describe “structures below”, therefore implying a vertical order between structures that can be considered foundations (and are usually invisible) and the layers above that are supported by this “underlying plumbing”. Today, the term infrastructure is usually associated with the process of development, for which infrastructures are considered to be a prerequisite. It has become a broad and dynamic term that is no longer reserved for the technical and physical artifacts underlying energy, railroad and telecommunications systems, as it can include the social and intangible systems our societies rely on, from education systems to law or even languages24. This contemporary understanding of the word can first be found in the economic development literature debates of the 1950s. According to William Rankin, it was after the Second World War that infrastructure became “a label for the technical-political systems required for growth and modernity”25. This economic literature has sought to rationalize the use of the term. Such rationalization was necessary to justify and normalize investment policies across national borders in the context of both decolonization and the emergence of official development assistance programmes. In this context, one of the key features of infrastructure is the provision of interoperability for the secure and stable circulation of people, goods, capital or data, by setting norms and standards. Railroad systems for instance need a standard track gauge, and languages need agreed grammatical rules to fulfill their purpose.

At the beginning of the twentieth century, infrastructure was still widely associated with the notion of 'social overhead capital'. It was considered a special kind of capital that is, in opposition with conventional capital, shared by several enterprises and can therefore not be attributed to a single productive activity26. In other words, infrastructures generate social benefits that exceed the private benefits of the mere production or use of an infrastructure by single enterprises. Indeed, infrastructures have so-called positive externalities (or spillover effects) that can affect third parties or even social and economic development in general27. Such effects are difficult to measure because the causal links between infrastructure and its benefits are mostly indirect. Infrastructures do indeed provide a generative input to other economic activities, which can be very diverse. You do not use an electricity network for its own sake, but in order to do something else. And it is very difficult to capture and compare all the different uses of a road: using a road to transport commercial goods is very different from using it to go to hospital. Some attempts to classify different types of infrastructure have been based on the typology of derivative uses to which they provide input, for example distinguishing between infrastructure used to produce market goods, public goods or non-market goods28. Nevertheless, fully understanding and modeling the impacts of infrastructure remains a challenge.

Attempts to normalize infrastructure have identified two key economic features: a demand side characterized by 'publicness' and a supply side characterized by “capitalness” (or capital intensity). The demand side is characterized by publicness because infrastructures are usually considered as public goods. This means that they are non-rival and non-exclusive: many people should be able to use the infrastructure at the same time and access to it is usually unrestricted or difficult to restrict. The supply side, on the other hand, is characterized by capitalness, as infrastructure tends to have very high fixed costs and relatively low operating costs29. For example, while the cost of installing and deploying a national telecommunications network is huge, the marginal cost of an additional user of the network is very low or even close to zero. This cost structure makes it difficult to set individual prices.

These two economic features have been used in the classical economic literature to explain why infrastructure can be considered a natural monopoly. Without the guarantee of a long-term economic rent, there is no incentive for self-interested actors to make the huge investments required to build the infrastructure in the first place, especially if individual access to the infrastructure or to the positive externalities infrastructure produces is difficult to restrict. This is seen as the origin of the challenge of providing public goods: collective action is needed to avoid free riding - where individuals benefit from the public good but have insufficient incentives to contribute to it. From a classical perspective, this specific market failure justifies government intervention. Indeed, one of the three roles of the state identified by Adam Smith - alongside protecting its population from invaders or from itself - was “that of erecting and maintaining those public institutions and those public works, which, though they may be in the highest degree advantageous to a great society, are, however, of such a nature, that the profit could never repay individuals, and which it therefore cannot be expected that any individual or small number of individuals should erect or maintain”30.

The interpretation of this economic division between public and private has varied over time. It has contributed to the "modern infrastructure ideal" of the second half of the twentieth century, characterized by state monopolies and tight control over infrastructure. It also provided the basis for the neoliberal turn that eventually led to increasing competitive private or public-private provision of infrastructure based on “selective premium infrastructural configurations”31 such as leasing deals to private companies or the commercialization of infrastructure services. From a legal perspective, the theory of essential facilities has provided the ground for a competitive provision of infrastructure. Essential facilities are indeed “an asset or infrastructure to which a third party needs access to offer its own product or service on a market” where “no reasonable alternatives are available and duplication of the facility is not feasible due to legal, economic or technical obstacles”32. Opening up traditional monopolies to competition came with increasing financialization of infrastructure. According to Kathryn Furlong, the increasing financial engineering of debt-led infrastructure development transformed the unattractive high investment in public works into assets that could be marketed to private investors. As these assets can be sold on financial markets dependent on global debt dynamics, they can even create a “disconnect between the returns to infrastructure and those to investors”33. This trend, which enables the generation and extraction of wealth for an increasing number of infrastructure projects, has contributed - in combination with deregulation and public austerity measures - to their continued privatization, which in turn threatens universal and equal access to them34.

Contemporary financialization shows the limits of classical economic definitions of infrastructure based on publicness and capitalness. This is in line with the social science literature that rejects infrastructure as a neutral economic concept and emphasizes its normative and political nature, to the point of considering infrastructure as the “setting and stake of social struggle”35. This makes infrastructure dependent on the specific socio-historical context in which it is embedded. It also means that infrastructure emerges from a political negotiation of what are perceived as “the most basic elements of collective consumption (and collective disposal) organizing social, economic and environmental lives which are lived, at least partly, in common”36. As Rankin shows, our economic understanding of infrastructure is still heavily influenced by the theory of staged growth and the post-war need of industrialized states to negotiate extraterritorial norms and material priorities for international trade and cooperation37. Such negotiations may formally take place at the level of governments, but they tend to “appear as complex points of intersection between state and speculative interests, and narratives of industrial restructuring”38. They can emerge outside clearly coordinated public efforts, especially in the context of rapid technological change: “against the will of their creators — and also against the concepts of politicians — infrastructures seem to have a will of their own. They are not only pushed through from above but also demanded from below, conflicting interests influence their shape, and they are used and rearranged in reaction to actual needs”39.

This reconfiguring nature of infrastructure has led to an increasingly abstract use of the notion, especially in the field of science and technology studies (STS). Infrastructure then becomes a specific set of relations - usually made of hybrid networks of humans and machines - that is dependent on a specific perspective. This perspective is a problematization of various yet fundamental aspects of collective life that allow for the trafficking of "goods, ideas, waste, power, people and finance"40. As such, infrastructure constitutes a methodological tool that can be mobilized strategically to create new knowledge and achieve certain political goals41. This literature analyzes contestations of “the technical procedures of expert judgment in infrastructure planning and management”, especially “through claims and counter-claims about the values produced by infrastructure, about the publics those values serve, about the kinds of expert or non-expert knowledge that are relevant for defining these values, and about the technical means required to realize them”42. Such contestations come for instance from political movements that have historically opposed infrastructure projects, criticizing their negative impact on the environment or on inequalities and wealth extraction, but also the specific threats in case of accidents or disruptions. Such movements oppose to “Big Infra” projects steered by government and corporate interests in the name of more distributed and resilient infrastructures, based on “auto-economies of people”43.

This recognition of the possibility of multiple forms of infrastructure at different scales, combined with the contemporary awareness of planetary boundaries and the "infrastructural gaps" left by the state and the market, has contributed to the recognition of new forms of collective responsibility for both physical and social infrastructures, and “posits that people are and should be the most integral part of infrastructure”44. In this context, infrastructure is often associated with the concept of the commons, as a resource management principle built on active involvement by the public45. It has also led to the emergence of a multidisciplinary field of research dedicated to “the often hidden work done in repair, custodianship, stewardship, tending and caring for the things that matter” as the founders of the “Festival of maintenance” put it46. An important distinction made in this literature focuses on the opposition between large-scale "hard infrastructure" projects, which are capital intensive, and local "soft infrastructure", which rely on sometimes fragile social structures and various forms of community participation for their maintenance47. This distinction already seems to indicate that commons represent a resource management principle better suited to the provision of infrastructure that is less capital intensive and more dependent on the ongoing contribution and maintenance efforts of people.

This first section has reviewed infrastructure studies spanning various fields and disciplines. It has shown that the concept of infrastructure always suggests a hierarchy between foundational structures, often invisible, and the layers above them. This hierarchy reflects the political negotiation of society’s basic elements of collective consumption and production. Frischmann defines infrastructures as “shared means for many ends”48, as they generate downstream economic activities and social benefits that exceed the private benefits of the mere production or use of an infrastructure by single entities. Infrastructures also enable secure and stable circulation of people, goods, capital, and data by setting norms and standards. They are characterized by their fragility, requiring significant financial investment and human effort for deployment and maintenance. Economically, infrastructures have indeed a demand side marked by "publicness" and a supply side marked by “capitalness.” Commons, as a resource management principle that has emerged as a more participatory alternative to the private and technocratic nature of “Big Infra” projects, appears to be better suited for infrastructures that rely more on ongoing human contributions and maintenance efforts than on heavy capital investment.

Competing definitions of the “public”

Infrastructure studies show that the economic perspective inherited from the development theory of the 1950s, which limited public intervention to market failures, is increasingly being challenged. A growing number of voices indeed call for a “distinctive role of public policy” that aims “not to boost private consumption by delivering economic growth but to ensure the quantity and quality of foundational services”49. This trend raises the question of redefining the private/public dichotomy, the nature of publicness and its ability to incorporate various forms of collectivity. The notion of commons is playing an increasing role in these debates. While social movements mobilize the concept to reject “deep structural categories embodied in the dualities of state/market, public/private, objective/subjective and universal/local”50, commons and digital commons in particular, are de facto already playing a key role in the provision of some infrastructure, such as the “digital infrastructure stack”51, as we will see in the next section.

The progressive abandonment of the private/public dichotomy inherited from development theory implies, according to Frischmann, to not exclusively focus on the “supply-side issues” of infrastructure, especially the problem of “securing cost recovery and incentives to invest in the face of decreasing-cost phenomena”. Instead, more attention needs to be given to the “tremendous societal demand for public infrastructures”, which can not be solely understood through the willingness of private user’s to pay for them in a market setting, as such an approach totally evacuates the large and positive spillover effects infrastructure can have on society52. Reformulating the demand for infrastructure as a collective demand that can’t be broken into aggregated private needs leads necessarily to a delimitation problem. In fact, if private actors are not the right unit to understand infrastructure demands, what is the right unit to represent collective interests? And what are the institutions able to represent such collective interests?

A first key feature of public infrastructures that most authors agree with is that they should be publicly accessible. The economic literature considers that goods that “naturally” present this feature should be considered public goods, justifying the involvement of the State in their management. But public access could also be understood as a normative feature of non-excludability, stemming from the Roman tradition of res publicae, which opposes both private and public property. It means that certain goods should not be owned exclusively by anyone, including governments, which contradicts the state ownership often implied in economic literature. This brings the concept closer to the idea of the public domain, goods from which no one should be legally excluded53. One can nevertheless consider that public infrastructures should in any case have “public attributes”, in the sense that they aim to provide universal and unrestricted access.

A second feature of public infrastructures is their participation in public goals. A first general understanding of this feature could be based on the opposition with private goals: infrastructure that would only serve one or a few entities’ interest or profit could therefore not be considered as public. The public function of infrastructure is nevertheless a highly contested one, as the notion of “public interest” covers different understandings depending on legal, historical, socio-cultural and political contexts. While the French understanding of public interest for instance historically referred to the interest of the State - as a body representing the nation as a whole - the notion of “common good” in the English context usually refers to a general interest of society that is not essentially the responsibility of the State54. This is reflected in the European Union’s definition of services of general interest, which acknowledges States’ capacity to define what these services are but also recognizes that they can be provided by non-public organizations as well. The EU recognizes the following three categories of general interest categories: services of general economic interest “which are basic services that are carried out in return for payment, such as postal services” ; non-economic services that include police and judiciary systems, and social services of general interest, which “respond to the needs of vulnerable citizens, and are based on the principles of solidarity and equal access”, such as “social security schemes, employment services and social housing”55. For David Eaves, Mariana Mazzucato and Beatriz Vasconcellos, similarly, public functions of infrastructure can be characterized by public goals such as fostering community and social relationships, fostering economic activity, guaranteeing better quality of life or guaranteeing essential capabilities56.

For these authors, however, the public attributes and functions of infrastructure are not enough to ensure the maximization of the public value of infrastructure, which needs to take into account “the processes surrounding value creation and maximization, and the political economy implications”. For infrastructures to maximize public value, “proactive governments who set the direction for the required collective action”57 are therefore required. This approach reflects a new trend in the management of infrastructure which goes beyond the classical public/private dichotomy and implies that “government direction, centrally defined public purpose, and large-scale planning are combined—in still-emergent ways—with market mechanisms, private actors, and public input”58. Similarly to Eaves et al., we will assess institutional arrangements to deliver infrastructure according to their capacity to maximize public value creation, accepting that the societies' views on what constitutes public value are diverse and can change over time. To be in accordance with the sociological critique that infrastructure too often rests on expert judgment only, we will not only assess if these arrangements can deliver infrastructure for the public (public interest), but also to what extent they can be delivered of and by the public (public ownership).

Public ownership extends the debate beyond the question of participatory governance models that provide a space for the public to be consulted, and includes questions regarding the active participation of the public in the control, funding and provision of infrastructure as well. While it is unrealistic for every individual to participate in all aspects of every infrastructure, several authors have proposed "commoning" as a form of "differentiated publicness."59 This concept promotes public participation through the involvement of civil society communities, either focused on specific territories or particular public issues. In this framework, public ownership extends beyond government ownership. Public provision of infrastructure could involve citizen participation or decentralized contributions from various entities, not solely relying on public servants. Public funding could encompass numerous small private donations. Public control might include various governance forms where citizens and civil society organizations have decision-making power, not only through their public representatives. Based on this view, Wikipedia could be considered a model of public infrastructure. Wikipedia has clearly public attributes and functions, as its content is shared through a creative commons license, ensuring non-exclusive use, and as it supports public goals, which are access to knowledge and education. Additionally, one could consider that is produced by the public, in this case by a large number of volunteering citizens, funded by the public, as the Wikimedia foundation relies on millions of small donations, and governed by the public, as the non-profit foundation hosting Wikipedia has established open, participative, and democratic decision-making processes to rule its management60.

However, delegating power to citizens or communities raises fundamental questions about infrastructure management. Public institutions are not only expected to be democratic but also to adhere to specific rules and regulations. When public services are delegated to the private sector, this typically comes with public service obligations aimed at mitigating potential negative externalities, ensuring inclusion (both equal access and equitable distribution of value), and providing effective transparency and accountability mechanisms, such as audit systems and grievance redress mechanisms. Additionally, these obligations are usually sector-specific. According to Mazzucato, there are two additional challenges in delegating public services and utilities. The first is ensuring "collective learning and building long-term capabilities and capacities" to counterbalance the strengthening of private expertise, which can create dependencies that negatively affect the public. Secondly, she argues that only states can ultimately ensure the overall "purpose and directionality" necessary for decisions impacting society as a whole61.

Table adapted from Davis Eaves et al.

This section has shown that contemporary perspectives on infrastructure management mark a shift from traditional economic views that confined public intervention to addressing market failures. There is a growing emphasis on public policy's role in ensuring the quality and accessibility of foundational services. This shift challenges the traditional private/public dichotomy and calls for integrating various forms of collectivity into public infrastructure. Public infrastructures should therefore not only have public attributes and functions but also a form of public ownership, which can range from ensuring public controls to public funding or even direct public provision. The concept of commons, particularly digital commons, is gaining prominence as it promotes public participation through the involvement of citizens, civil society, and private entities in resource management. However, this approach also raises questions about how to represent collective interests effectively and ensure the accountable and transparent management of infrastructures. In the following sections of this paper, we will investigate strategies to maximize public value creation in contemporary debates on digital public infrastructure. We will examine how these strategies serve the public interest and how they involve the public in the ownership, funding, and control of digital infrastructures.

The potential spheres behind public digital infrastructure

This section provides an overview of the policy debates on public digital infrastructure, focusing on policy areas where digital commons already play a critical role or where public institutions are developing strategies that include digital commons as potential infrastructure providers. It will briefly review some of the history of digital infrastructure, showing how the free market policies of the 1990s have led to the current platform-driven digital environment. In this context, digital commons are often perceived as an alternative that could participate in the development of a protocol-based digital landscape relying on collective governance and non-extractive economic models. It is important to note that debates about digital commons as providers of public digital infrastructure tend to focus on infrastructures that are purely digital (made of bits), excluding the physical layers of infrastructure formed by network and computing infrastructures. This observation confirms the assumption made earlier that commons are appropriate providers of public infrastructure when the supply of that infrastructure relies more on human contributions and maintenance efforts than on large capital investments.

Telecommunication networks, and to some extent even the first computer technologies and the early Internet, were born in the era of the "modern infrastructure ideal". Postal, telegraph and telephone services were historically provided by publicly regulated monopolies or even directly by governments. In the 1960s, the concept of the "computing utility" initially envisaged shared computing power similar to these telecommunications services, ensuring accessibility and affordability for all. Analysts were imagining “giant computers whose central processing unit (CPU) time could be shared, in the same way that electric utility customers share huge power plants”. The French Minitel system introduced in the 1980s was based on similar premises62. Across the globe, governments played a significant role in developing foundational technologies, as exemplified by the early development of the Internet sponsored by government agencies such as the US Department of Defense's Defense Advanced Research Projects Agency (DARPA) or the US National Science Foundation, who laid the groundwork for a shared network accessible to researchers and, eventually, the general public. Economist Mariana Mazzucato has described this government involvement in innovation and technological change as the hallmark of an "entrepreneurial state", which she observes across a wide range of key technologies - from essential components such as the hard disk drive, microprocessors, memory chips and LCD displays, to the development of GPS, the HTTP protocol, touchscreen technology or even virtual assistants63.

The internet in particular, as an ubiquitous network of networks, based on the globally adopted TCP/IP protocols, has been considered as a key contemporary infrastructure, especially because of the many essential products and services that depend on it today. It’s public governance model, based on multi-stakeholder participation, has conducted many researchers to consider it both an infrastructure and a global commons64. The actual provision of the internet was also largely funded by governments, who invested in the “information superhighways” to support the public uptake of a network originally still mostly used by academics. However, with the advent of neoliberal policies, internet service provision became largely deregulated and fragmented into private companies. Today, most internet service providers are private companies, with a few countries offering state-sponsored access to connectivity in public spaces, and only a handful of community-owned networks managed as digital commons resisting65. While the physical networks managed by telecom operators are still considered infrastructure and regulated accordingly by policymakers, this model has consistently posed challenges to providing universal, affordable and reliable access, especially in remote areas66.

The development of services on top of this infrastructure was largely left to market powers, resulting in an unprecedentedly rapid accumulation of private wealth, with some of the most highly valued companies in history emerging in the digital economy. A large number of analyses have provided explanations for these processes, among which the large network effects that have led to “winner-takes-it-all” situations, data extraction, or the technical possibility to lock-in users within a product environment or an ecosystem of services. These strategies have been mobilized by contemporary platforms to become bottlenecks of the digital economy - or even essential facilities of the economy in general, notably through their “exclusive control over search engines, ecommerce platforms, and app-stores”67. While platforms share many features with infrastructure, considering that they serve as a beneficial foundation for a variety of uses and have become truly unavoidable, they “leverage programmability and interconnection to achieve control, rather than relying on direct provision and expansion”68. This ability to coordinate heterogenous and independent actors while extracting value from these interconnections has allowed platforms to challenge existing traditional infrastructure, buying the submarine cables forming the internet backbone, competing with broadcast and media infrastructures, or combining cloud computing power and software components, merging them into business models presented as “infrastructure as a service”69. The vertical integration strategy pursued by these platforms, combined with recent technological developments, in particular the extensive development and rationalization of centralized cloud computing architectures, best exemplified by the systemic power of massive-scale data center providers (hyperscalers), and the development of artificial intelligence (AI), especially the global hype around generative AI models, has given these platforms additional infrastructural power. Concentrations in cloud computing by companies such as AWS, Microsoft Azure, and Google are already creating a bottleneck for entire cloud computing market segments and threaten to create an additional one over AI systems and applications in the future70.

This increasing private concentration of infrastructural power has led to numerous policy debates on digital sovereignty and the need for renewed industrial policies, especially in Europe. But because of the amounts of capital necessary to compete with existing private infrastructures, European countries have still not found any public and coordinated answer to the current situation71: in the first half of the year 2024 alone, Microsoft has announced that it would invest 4 billion euros in France72 and 3,2 billion euros in Germany73 to further develop its cloud and AI infrastructure.

For the same reason, digital commons do not seem to be suitable as alternative infrastructure providers in this area. Indeed, community networks and community hosting services are currently clearly limited to providing local, small-scale and grassroots alternatives, struggling to raise the necessary capital and to adapt to public service obligations and regulations designed for private “Big Infra” projects74. Digital commons do, however, offer alternatives on the immaterial layers of digital infrastructure from which most private platforms have emerged, notably because of the scalability of non-rival digital resources. The platform model indeed contrasts with the protocol-based network of networks that characterizes the Internet. Both create the conditions for interoperability, allowing people to connect and exchange data, ideas, goods and services, but the platform model is based on an architecture of centralized control and value extraction. While the platform model has inspired many other organizations, especially in the context of the increasing rationalization and externalization of economic activity, leading to a degree of 'platformization' of infrastructure75, their unprecedented power and wealth has also led to counter-claims that see public digital infrastructure as the very opposite model to platforms. From this perspective, public digital infrastructure is redefined as “a shared set of rules and open protocols”, where no single actor “owns a whole suite of tools and can unilaterally set the rules”. Such infrastructure would lay the ground for “a protocol-based economy”76 producing “generative interoperability”77, that would, unlike platforms, allow for a variety of downstream uses without centrally controlling or extracting value from them. Public digital infrastructure therefore comes with a renewed role for the public, which requires addressing both a vast array of socio-economic development challenges and the current concentration of digital markets78. The European Union, for instance, has taken many measures to develop legal environments that contribute to an increased interoperability but also established strategies to be more present in international bodies where technical norms and standards are established79.

In the context of these trends, the rest of this section will examine current debates on public digital infrastructure - meaning where alternative architectures to the private platform model are discussed. It will be limited to the infrastructures that are purely digital (made of bits), excluding physical layers on the one hand and immaterial and social layers on the other, while acknowledging the intertwined nature of these systems. As mentioned in the previous section, we will consider infrastructures as “shared means to many ends”, resulting from the political negotiation of society’s basic elements of collective consumption and production. The paper identified the following four areas of policy debates where social demands for public digital infrastructure can be clearly identified, and where digital commons can be part of strategies to maximize public value creation:

• Common foundations of government solutions for public service delivery,

• Open source software components for the Internet stack,

• Critical layers for access to digital public spaces,

• Platforms and services that are critical in specific sectors for industry or intermediation between producers and consumers.

E-Government foundations as Digital Public Infrastructure

The discourse on digital public infrastructure (DPI) has garnered significant attention, particularly due to the pioneering efforts and advocacy of the Indian government. India has conceptualized DPI as open and interoperable technologies that facilitate essential functions for both public and private service delivery80. This concept is notably inspired by India’s advancements in digital identity, e-commerce, and payment systems. India's DPI model, known as India Stack, comprises a set of open application programming interfaces (APIs) considered "government-owned, non-competing goods... made available at low costs"81. This model aims to create an ecosystem of services through DPIs: public institutions primarily offer networks or gateways, on top of which independent but interoperable services can be developed by individual entities. India’s DPIs were mostly developed through public-private partnerships that closely associated India’s non-profit organization iSPIRIT, which represents the Indian software industry82.

A key example is Aadhaar, launched in 2010, which is the world's largest biometric ID system. Its open architecture enables various sectoral applications, making it a foundational system for accessing public and private services. However, it has faced criticism regarding security and privacy risks associated with its centralized nature. Additionally, the public platformization strategy adopted as part of the India Stack development has been questioned because of its increasing involvement of for-profit entities in welfare delivery, healthcare, and education. According to Eshani Vaidya, the fact that private entities play a significant role in building and managing digital services can lead to a lack of accountability and contribute to social exclusion, particularly for marginalized groups83. Some of these criticisms have led to the establishment of more robust regulatory frameworks around the adoption of DPI in India, and triggered international debates on the need to establish safeguards before deploying DPIs84.

As noted by Vy Dang et al., the India Stack has nevertheless been lauded for fostering innovation in both the private and public sectors. It is seen as a significant driver of economic growth, with projects like Aadhaar and India’s payment system UPI contributing to 0.9% of India’s GDP in 202285. The public sector's use of the India Stack has also allowed it to streamline key government services, notably during the pandemic. For instance, the Co-WIN digital infrastructure facilitated the administration of over 2.2 billion COVID-19 vaccines nationwide86. The importance of DPI to address common governmental challenges and accelerate development is now endorsed by a range of actors that reach far beyond India.

Estonia’s X-Road, an open-source government data exchange system, is an example of a European DPI often mentioned internationally. The scalability and adaptability of these solutions have attracted significant interest from various stakeholders, including development organizations, consultancy firms, and digital vendors, eager to leverage these experiences to deliver new services to governments. Organizations that have adopted India’s vision of DPI include the United Nations Development Programme (UNDP), the G20 and private sector giants like Huawei, Amazon Web Services, and Mastercard87.

Open source software as critical infrastructure of the Internet Stack

Another important debate on digital infrastructure concerns more directly the development, maintenance, and security of the various collectively shared layers of the Internet stack. The boundaries of this stack are difficult to establish because of the growing interconnection and interoperability required among all information systems and devices. As put by Berlinguer, “it is not just physical infrastructure, such as cables, web servers, hardware or data centers, but also infrastructure made, for example, of software, protocols, data, standards, operating systems, and programming languages. (...) The increasingly pervasive and ubiquitous intermediation of digital networks and data flows (...) is contributing to potentially expanding and blurring the scope of the notion of infrastructure”88. The analogy of the stack has been recognized to account for the complexity and stratification of these systems, giving the possibility to introduce multiple perspectives while showing interdependencies and hierarchies89. Because many of the shared layers of the Internet stack are based on open source software (OSS) and are governed collectively, they have often been described analytically as digital commons90,91, which together form an “accidental megastructure" without any unique and central form of coordination.

The growing political focus on various Internet stack layers is mostly the result of a realization of the critical and global role of OSS, with 96% of codebases containing OSS92 and “roughly 70-90% of any software stack” consisting of OSS93. In this context, the foundational role of OSS for a vast array of downstream activities has been widely documented. Frank Nagle estimates that without OSS, companies would face $8.8 trillion in software development costs94. A significant study commissioned by the European Commission also highlighted the economic value of OSS: the research revealed a cost-benefit ratio of 1:4 for OSS investments and demonstrated that OSS substantially contributes to the EU's GDP, with a 10% increase in OSS contributions potentially boosting the EU's GDP by 0.4% to 0.6% annually95. Many open source tools are indeed crucial in driving scientific and technological progress, aiding breakthroughs in life sciences, physical sciences, and the digital economy: the open source tool NumFocus allowed for instance to produce the first image of a black hole96. Additionally, OSS can create shared industrial goods vital for sectors like the automobile industry97 and support the growth of SMEs98, supporting the case for public investments into OSS as part of an “entrepreneurial State” strategy that would proactively support foundational innovations. Digital commons are also increasingly mentioned in policy debates on alternative ecosystems and value chains in the field of AI99, based on a mutualized access to critical resources such as datasets and governance models that ensure these resources are resilient to extraction or capture by private interests100.

But the interest in the OSS infrastructure behind the Internet stack is also driven by several external factors: heightened scrutiny of value chains and global dependencies due to the Covid-19 pandemic, rising geopolitical tensions, technological competition, information warfare, and the militarization of digital infrastructures101, contributing to the popularity of proposals aimed at strengthening States' digital sovereignty. In this context, some vulnerabilities of OSS such as the OpenSSL case have generated political attention to a still widely politically underinvested field. In the 2016 report “Roads and Bridges: The Unseen Labor Behind Our Digital Infrastructure”, the risks of contemporary overreliance on underfunded infrastructure, largely supported by volunteers driven by reputation-building, obligation, or passion, has been established102. More and more attention is being paid to the maintenance required by digital technologies, which has historically been associated only with innovation. In the context of OSS, maintenance is essential to keep software up to date and compatible with new systems, to fix bugs, or to patch security vulnerabilities. The volunteers who perform these activities - which tend to become more complex and necessary over time - are increasingly overwhelmed and become potential targets for social engineering attacks, as in the case of the recent XZ utils vulnerability103. Many analysts see this as a prime example of the "tragedy of the digital commons"104, where many benefit from a shared resource without contributing back to it in an appropriate way, such as by donating to an open source project's host, who is typically responsible for its maintenance. In response, several calls - both from academia105,106 and open source communities themselves107,108 - have been made to consider OSS as part of our public infrastructure, and therefore as a resource that should receive public funding.

In this context, the suggested role of the public is to contribute to the overall resilience and security of some “critical infrastructure” components of the Internet Stack. This approach typically begins with developing internal intelligence and understanding of critical systems and their dependencies on societies, economies, and public institutions to identify systemic digital risks. A notable example is the “Free and Open Source Solutions for European Public Services” (FOSSEPS) pilot initiated by the European Parliament to map the common dependencies of European institutions on OSS109. Unlike the traditional tendency of states to focus inwardly on their largest national industrial players, this approach aims to create a more decentralized digital landscape by developing funds that can provide sustainable and long-lasting support for the OSS ecosystem110. The suggested strategies also involve opportunistic investments in targeted packages, by tracking dependencies and vulnerabilities, conducting risk assessments, and identifying needs from the grassroots level. This requires ongoing dialogue between OSS communities and dedicated public institutions, exemplified by the German Sovereign Tech Fund, an initiative launched in 2022111. These ideas are also reflected in the industry-inspired trend to establish Open Source Program Offices (OSPOs), in order to mainstream collaboration with OSS communities in public administrations112.

A public stack to access Digital Public Spaces

Another important debate on public digital infrastructure has emerged in response to the private-led development of digital services in the 21st century, which has - as demonstrated above - resulted in the dominance of a few large companies over key infrastructures not only for the economy but for society as a whole. These debates focus on the risks to human and digital rights posed by the provision of key digital services - especially widely used - by monopolistic players whose business models are based on surveillance capitalism113. While — on the level of the EU — some of these risks have been addressed by a new set of regulations that emerged in 2010 as a response, including the GDPR, but also the DMA and the DSA, many authors still argue that public institutions should play a stronger role in creating non-private alternatives for online speech and debate, which includes many social media platforms, forming spaces for political and civic engagement, spaces of knowledge sharing and scientific production, or educational and cultural spheres.

The 2020 Waag Report on “Online European Public Spaces”, for example, argues that while physical public spaces are governed by a social contract that protects public values, digital spaces lack such basic agreements. Social media platforms are often perceived as digital public spaces due to their widespread accessibility and free use, but they still operate under private rules without democratic accountability. In fact, platforms enforce their own rules and use opaque algorithms to manage content, curating political discourse based on hidden rules and agendas114. Given that a digital world dominated by private corporations cannot effectively guarantee privacy, democratic debate, and human rights, there is a growing social demand for “digital services and platforms that exist outside the control of commercial entities that extract value from users of their platforms”. According to Paul Keller and Zuzanna Warso, public institutions should therefore participate in the design of “fora for public and private exchanges, access to information, and tools for civic organization” that “adhere to democratic and collective forms of governance”115. This underlying vision for “Digital Public Spaces” has been adopted by several European stakeholders in a Manifesto calling governments to take more responsibility in the maintenance of a “public stack” for a “Shared Digital European Public Sphere (SDEPS)”116.

In his definition of digital public spheres inspired by Habermas, Christian Fuchs emphasizes on the “democratic, non-capitalist and unideological character” of the public sphere in order to effectively facilitate critical public debate. In this context, the technical infrastructure providing access to such spheres should be “publicly owned and managed by independent, non-profit organizations''117. This approach is similar to the research results and policy recommendations of Ethan Zuckermann’s Initiative for Digital Public Infrastructure at the University of Massachusetts at Amherst118. For Zuckermann, governments should support new online media institutions, similar to their establishment of public service broadcasters such as the British Broadcasting Corporation (BBC) at the beginning of the 20th century. Such online media institutions should be “publicly spirited”, but diverse in funding”, “plural in purpose”, “participatory in governance”, and “publicly auditable and reviewable”119. Several proposals to reinvent pluralism in the digital age have adopted the idea that public institutions’ role should be to allow citizens to exercise their rights by creating independent public spaces where citizens could individually and collectively decide about moderation rules, for instance by developing alternative forms of algorithmic recommendations120.

Many authors - from Zuckermann to Fuchs - consider Wikipedia and other digital commons to be a good illustration of the new type of public service media that governments should support. For Keller and Warso, digital commons are also “a key mechanism for the provision of a public digital infrastructure”, while recognizing the necessity of significant public intervention to enable non-profit alternatives to scale and compete effectively121. This raises the question of defining a new relationship between public institutions and digital commons. In addition to the historical European models for investing in media landscapes to ensure plurality, contemporary developments in knowledge infrastructures can offer interesting insights in this respect. In the context of the advent of open science, many public institutions have indeed financially supported the development of open infrastructures which are largely hosted by non-profit organizations, with “strong and widespread commitment to community engagement”122. For Mounier-Primbault, the example of knowledge infrastructures demonstrates that future infrastructures should not only rely on “supposed non-values (such as rationality, efficiency, emergency, short-term)”, but also “embed values” to offer an alternative model to the extractivism of the platform mode and be part of a resilient and sustainable community ecosystem123.

Collectively-owned platforms to organize production

A final important area of policy debate on public digital infrastructure and alternative architectures to the private platform model is represented by debates inspired by the platform cooperative movement. Platform cooperatives are collectively owned and democratically governed, in contrast to venture capital-funded platforms. The term "platform cooperativism" was introduced by Trebor Scholz in his 2014 article, "Platform Cooperativism vs. the Sharing Economy."124 The goal of his article was to shed light on the exploitation and manipulation of both workers and users by platforms, and to suggest that both groups could benefit from alternative models. The term is now largely associated with solutions to the abuses of the "gig economy," which involves an increasing reliance on temporary and part-time work performed by independent contractors and freelancers, but also significant transfers of wealth as commissions and fees are applied by these international platforms to sometimes very local transactions. According to the EU, “over 28 million people in the EU work through one (or more) of these digital labor platforms.​ In 2025,​ that number is expected to reach 43 million people”​125. While various regulatory efforts have been made to mitigate some of the negative externalities and impacts of platform work, such as the Platform Work Directive, several actors, including the Platform Coop Consortium, are calling for more active support from public institutions for the development of collectively-owned platforms. Such policies are considered to offer opportunities to restructure current value chains for digital labor and to contribute to local economies’ development126.

Collectively-owned platforms include various alternatives that can range from platforms established and owned by public institutions to platforms managed by informal collectives. The report "Policies for Cooperative Ownership in the Digital Economy" by the Platform Coop Consortium and the Berggruen Institute examines government policies affecting collectively-owned platforms in various regions. The report covers several supportive measures that have already been implemented, such as cooperative legislation reforms, tax benefits, public procurement incentives, business support services, and cross-border municipal cooperation. While these policies have fostered a conducive environment for some platform cooperatives, and while examples of publicly-owned platforms for local transportation exist in certain cities, for instance in Brazil, the report acknowledges that no collectively-owned platform can yet compete with venture capital-funded platforms and their financial resources. The report therefore concludes that the public funding of cooperative platforms should be part of “national, regional, and municipal development strategies” that blend regulatory measures with funding that could also include “direct state ownership”127, an approach which is reminiscent of approaches advocating a form of industrial policy.

A summary of policy debates on public digital infrastructure

In reviewing current policy debates on public digital infrastructure, several key issues have emerged. First, infrastructure policies are the results of a specific and political problematization of social relations, that defines a hierarchy between elements that are perceived as foundational because they provide a generative input into a wide range of activities. This makes infrastructures recognized for their society-wide economic and social functions, even though their overall impact and spillover effects remain challenging to measure. Secondly, from a supply perspective, infrastructures can require substantial initial investments. Their inherent fragility and potential need for reconfiguration demand ongoing maintenance, which involves significant human and financial resources. Commons, as a resource management principle that has emerged as a more participatory alternative to the private and technocratic nature of “Big Infra” projects, appears to be better suited for infrastructures that rely more on ongoing human contributions and maintenance efforts than on heavy capital investment. This observation also applies to digital infrastructures, where digital commons struggle to offer alternatives in the capital intensive and physical physical layers of infrastructure formed by network and computing infrastructures, but seem to participate in the development of a protocol-based digital landscape relying on collective governance and non-extractive economic models.

Infrastructures must ensure interoperability for the secure and stable circulation of people, goods, resources, capital or information. They possess the capacity to set norms and standards, but also to include and exclude. For this reason, effective accountability mechanisms, such as audit systems and grievance redressal mechanisms, are essential components for public infrastructures. The public nature of infrastructures is subject to different interpretations. Modern understandings do not limit this nature to government ownership, and rather analyze publicness as a spectrum or as a process that should maximize their public value. Hence, while public infrastructures should have public functions and attributes, the degree of public ownership or public involvement in the production, funding and control of infrastructures can vary.

While the free market policies of the 1990s have led to the current platform-driven digital environment, the high social demand for public digital infrastructure has led to new policy debates that position the state as an “entrepreneurial state”, which provides infrastructure that is built on the principles of openness - favoring a digital landscape based on “generative interoperability”. Policy debates focus on the diverse public functions that digital infrastructures should support and the derivative uses they should enable. E-governance foundations, the critical components of the internet stack, public spaces, and shared production platforms are seen as vital generative inputs necessary for contemporary economic and social well-being. There is a consensus on the public attributes of the technical objects forming public digital infrastructure, particularly their openness and non-exclusive nature, which qualifies most of them as digital public goods. Various forms of public ownership have been proposed for these different spheres of public digital infrastructure, reflecting a nuanced approach to ensuring their public character and maximizing their societal benefits.

Digital commons play an important role in the provision and management of many of the alternatives to the private platform model and can play a role in the public provision of digital infrastructure in all the fields currently discussed by policy makers. For e-governance foundations, the strategy usually involves public institutions controlling protocols and norms to ensure interoperability, supported by public-private partnerships. The participation of the development and maintenance of the Internet stack relies on an ecosystem strategy, with a mix of private and public funds to back open-source software (OSS) communities, with OSS projects managed as digital commons. In the realm of public spaces, public-commons partnerships seem to be key, with public funding supporting independent infrastructures managed as digital commons. For co-production platforms, the strategy deployed by policy makers blend policy instruments to restructure digital value chains, advocating for collective management of these platforms.

Digital Commons and the State

Characteristics of digital commons

In Section 1, we have shown the role that digital commons already play in providing some of the fields of public digital infrastructure that are currently discussed by policy makers and that their value is increasingly recognized in this context. This section provides an overview of the different ways in which the relationship between digital commons, the state, and public institutions has been conceptualized and theorized to date. In addition, we provide case studies to illustrate how the interplay has unfolded in different contexts, shedding light on the dynamics and implications for both digital commons and state actors. This analysis seeks to improve understanding of the relationship between the digital commons and the state, as the interaction between these entities becomes particularly important when digital commons serve as public digital infrastructure.

A substantial body of literature explores the concept of digital commons, and it is beyond the scope of this paper to offer a comprehensive review of the various stances adopted over the past few decades on this topic. For the sake of this publication, we assume that digital commons are distinguished by three key features. First, they can be decomposed into bits or binary digits, which not only makes them digital, but also makes them non-rivalrous. The second characteristic is community – digital commons are predicated on distributed production and are managed collectively by a group rather than by a single individual or entity. These communities coordinate not through pricing or subordination but through voluntary peer production. Lastly, digital commons are defined by a governance system with established rules for access and sharing of the resource. These rules are primarily designed to safeguard the resource’s development and sustainability against exclusive uses and profit-making and can make a resource completely non-excludable for instance when open-source licenses are employed. Moreover, the governance systems encourage deliberation and free participation by community members, ensuring their design promotes sustainable and inclusive use.128

Digital commons as the third way beyond the market and the state?

Much of the classic literature on the digital commons examines its role within the modern market economy and its potential to influence and disrupt market structures by fostering peer-to-peer collaboration and decentralized production. Benkler's work has been instrumental in characterizing the role of information commons and decentralized collaboration for innovation, information production, and freedom in the networked economy and society.129

The (digital) commons were sometimes seen as offering a radical possibility of a world beyond the market and the state130 and as a “third way of organizing society and the economy that differs from both market-based approaches with their orientation toward prices, and from bureaucratic forms of organization with their orientation toward hierarchies and commands;” a “socially progressive alternative to producing and sharing resources and to organizing collective action (...).131 In that sense, digital commons were considered as having "post-capitalist potential.”132

When (digital) commons were conceptualized as a counterpoint to both the capitalist market and the “centralized-controlled state politics”133, it positioned the digital commons and the state as opposing forces. Even, however, within this perspective of the digital commons-state relationship, the value and necessity of state support for digital commons, for example in areas like open science, is recognized.

In his analysis of FOSS, Berlinger highlights the evolution of commons studies, emphasizing the shift from viewing commons as autonomous spheres separate from the market and the state to understanding them as entities that interact with markets.134 He proposes three hybrid arrangements for understanding this interplay: semi-commons, shared infrastructure, and ecosystem creation.135 While the transition to FOSS is well underway in the marketplace, public administration and policymakers are struggling to effectively engage with the model. Berlinger attributes this to two factors. First, neoliberal political orientations discourage public policy and the public sector from directing and guiding technological development. Second, the individualistic and libertarian ethos of early FOSS communities further complicated the role of government and public policy in these environments. These factors make it difficult to rethink and redefine the role of government in FOSS-driven public digital infrastructure.

This analysis points to the fact that the relationship between the digital commons and the state is porous, and that neither the category of digital commons nor the role of the state in relation to them is monolithic, but can vary significantly depending on the political, economic, and social context.

Digital commons and public administrations

As noted above, the discourse of the commons is marked by a significant tension between the commons and the state. Many researchers and practitioners, particularly in digital communities that are more open to libertarian ideas, emphasized the decentralized and self-regulating nature of the commons, and have been skeptical of public intervention. At the same time, other researchers have seen the commons as an opportunity to rejuvenate or reclaim historic public services.

Traditionally, the relationship between the state and the commons has been conceptualized as a conflict between two distinct social systems - communal resource management (the commons) and state-driven processes such as privatization and enclosure. In the case of the commons, resources are managed collectively by a community, with rules and norms established by the community members themselves. On the other hand, the state, especially in its neoliberal form, often promotes individual ownership and market-based resource allocation. This has led to processes where common lands or resources are converted into private property. Such state-driven processes have historically resulted in significant social and economic changes, often at the expense of communal resource management.136 The most evident and widely quoted instance of this tension was the enclosures and privatization of common lands during the agricultural revolution in Europe.

The antagonism and tension between the commons and the state is not only historical but continues to manifest itself in contemporary contexts, including the digital realm. With the advent of the internet, new forms of commons have arisen, and the relationship between these digital commons and the state has continued to be a subject of theoretical inquiry. In the digital commons sphere this relationship has a distinct aspect related to the origins of the internet, which was initially propelled by government military funding. Despite this state-led initiative, the academic community played a significant role in the creation of the Internet and enjoyed considerable freedom in its development.137 Thus, the emergence of the Internet points to a deeply rooted interplay between top-down and bottom-up efforts to build and sustain the digital commons.

In the introductory chapter of “The Wealth of Networks”, Benkler points to the “rise of individual and cooperative private action and the relative decrease in the dominance of market based and proprietary action”. Then he goes on to ask, “Where in all this is the state”? Benkler notes that in both the United States and Europe the state has often supported market-based industrial incumbents at the expense of individuals in the emerging networked information economy. Benkler criticizes state interventions that cater to incumbents or attempt to optimize outdated modes of information and cultural production. However, he does not object to the state pursuing liberal projects and commitments and notes that his position is not rooted in a theoretical skepticism about the state. He suggests that the state could play constructive roles, such as funding neutral broadband networks and basic research, and regulating to prevent monopoly control over digital resources.

According to Benkler, the state’s role is limited due to the trajectory of markets and the rise of individual and social action in the digitally networked information environment. He sees nonmarket individual and social action as the most important domain for advancing liberal commitments, given the economics of computation, communications, information, knowledge, and cultural production. Despite this, Benkler does not resist many of the roles traditionally played by the liberal state. He concludes that once the networked information economy stabilizes and the importance of voluntary private action outside of markets is understood, the state can adjust its policies to facilitate nonmarket action and use its outputs to support liberal commitments. In other words, for Benkler, the state has historically disrupted and interfered with rather than supported the commons, but there is potential for a more nurturing relationship. At the same time, this potential is limited by the economics of commons based peer production.

The work of legal scholars such as Lawrence Lessig and James Boyle has emphasized the key role of state-adopted legal and institutional frameworks, particularly intellectual property law, in governing the digital commons and ensuring equitable access. More recently, other scholars have also pointed to the need for legal frameworks to protect the digital commons from enclosure or appropriation, given the risk of public investment being captured to produce open resources that are then commoditized by large corporations.138

Scholarship has also looked at legislative interventions that go beyond guaranteeing access and protecting the commons from enclosure, but that can "act as leverage for commoners.”139 Based on the analysis of five specific pieces of legislation adopted in France between 2014 and 2016, Peugeot suggested a typology of the legal devices protecting or promoting the commons. She distinguished four functions that encompass the law’s various capacities in relation to the commons: the ability to facilitate, protect, institute, and contribute. In this framework, facilitation involves the extension of traditional property rights structures, achieved through exceptions in copyright laws or the establishment of alternative management systems. Protection focuses on protecting the commons from the threat of enclosure by legally recognizing their existence or prohibiting exploitative practices. These two functions correspond to the roles of legal regulation recognized above. Institution, on the other hand, goes further in that it involves more active promotion of cooperative and democratic governance frameworks. Finally, contribution supports the commons either directly by providing open data or indirectly by adopting free software. Similarly, in the context of FOSS, Berlinguer looked at different pro-FOSS policy areas that ranged from allowing the use of FOSS in public administration, requiring that public procurement give equal consideration to FOSS alternatives, policies favoring FOSS solutions over proprietary ones, to attempts to make the use of FOSS mandatory in public administration. In the context of the creation and maintenance of scientific data commons, Contreras distinguished nine functional roles that state actors can play.140 He proposed the following categories: creator, funder, convener, collaborator, endorser, curator, regulator, enforcer and consumer, while recognizing that some of them are overlapping.

Some of the most recent studies on the relationship between the state (understood here as the public administration) and the digital commons by Sebastien Shulz look into the “commonization of digital public goods and services.”141 This is a process of blending (“hybridization”) of self-governed citizen communities with hierarchical public administration. This field of inquiry is guided by three main questions. First, how can equal access rights and administrative protection of sensitive data coexist. Second, how can the horizontal logic of peer production be integrated into the hierarchical production of official databases and software by bureaucratic administrations. And third, how can hierarchical management of public goods and services be reconciled with citizen self-governance institutions. These three questions correspond to the features of the digital commons: equal access, co-production, and self-governance. In the process of hybridization between digital common and public administration, these features are operationalized through open data initiatives, citizen crowdsourcing, and co-governance.

Shulz focuses on the power dynamics between citizens and administrators in co-governing digital public goods and services. Case studies analyzed by Shulz confirm there is a tension between the desire of citizen communities to self-organize and the hierarchical administrative procedures imposed by the state actors. Shulz identifies five factors that enhance and two that hinder citizen power in the co-governance of commonized digital public goods and services. Factors that enhance citizen power include

1. support within the administration for the commonization,

2. political context favorable to commons-like reforms,

3. willingness to adapt the existing state-centric institutions,

4. opening of working and organizational digital workspaces, and

5. the ability to modify or create new legal administrative rules so at to institutionalize citizen-centric co-governance.

The fact that only a small proportion of citizens are involved in shared governance and the tendency towards "state-ization" are, on the other hand, factors that limit the power of citizens in digital co-governance arrangements. While in this study, Shulz looks into commonized public goods, and this process is distinct from converting digital commons into public goods or public digital infrastructure, the factors identified by Shulz remain relevant in both scenarios.

Insights from the policy debates on digital commons and the state

Experts have emphasized the necessity for legal structures to safeguard the digital commons from being enclosed or appropriated. More recent studies have also examined legislative measures that not only ensure access and shield the digital commons from enclosure, but also serve as a catalyst for the digital commons.

In his study of the commonization of public goods, Shulz identified factors that enhance citizen power in the co-governance of commonized digital public goods and services. These factors are also relevant for enabling digital commons to serve as providers of public digital infrastructure. Essentially, a political environment that is conducive to commons-based reforms is a prerequisite for allowing digital commons to fulfill the functions of PDI. Political will and supportive policies that cover the different capacities of the state in relation to the digital commons (the ability to facilitate, protect, institute and contribute) must drive the transition and ensure its sustainability. Moreover, when considering the assignment of the role of Public Digital Infrastructure to the digital commons, public and state institutions may need to adapt to include elements of community management and collective decision-making. The ability to modify or create new administrative rules is critical for establishing the new arrangements and ensuring their legality and sustainability.

While the recognition of digital commons as providers of public digital infrastructure can bring some opportunities, it also has implications for their governance models. First, there's the risk of "state-ization," which refers to the potential encroachment of state control or influence over the digital commons. While increased state involvement could bring more resources and support, it could also undermine the autonomy and self-governance that are fundamental characteristics of digital commons. Government-imposed regulations or standards could stifle the community, changing the very nature and “spirit” of the commons. Second, serving as a public digital infrastructure would bring additional obligations and requirements. Digital commons would need to ensure a level of reliability, accessibility, and security expected of public infrastructure. This may require the implementation of more robust systems and protocols, which may require additional resources and expertise. In addition, increased public scrutiny and accountability would be inevitable given the public function that digital commons would serve.

Recognizing digital commons as providers of public digital infrastructure also requires a shift in policy. This shift must take into account the significance of the role that digital commons play and the different functions that the government can fulfill in relation to them. While the state's roles of facilitator, protector, “institutor”, and contributor already encompass a wide range of actions and responsibilities toward the commons, the recognition of digital commons as providers of public digital infrastructure introduces additional complexities and new responsibilities, in particular in terms of the need to maintain and sustain the commons.

In this new context digital commons may be subject to increased regulation and oversight. Policies would need to be developed to ensure that they meet certain standards of reliability, accessibility, and security. Policies would also need to be implemented to provide funding and resources to digital commons in recognition of their role in providing public infrastructure. These additional responsibilities could be interpreted as a new role for the state. In this role, the state assumes a more active and ongoing responsibility for ensuring the health, sustainability, and public utility of the digital commons. This includes not only protecting and contributing to the commons, but also actively managing and overseeing their operation in a way that serves the public interest.

As far as implications for public policy are concerned, finding from Berlinger's research on FOSS is also relevant.142 Berlinger emphasizes the need for a new generation of public policy to navigate the new political economy, which includes combining different regimes of ownership, governance, and value creation, as well as managing a tripartite governance system that consists of governments, markets, and communities. Policymakers need to address the shortcomings of each governance system. To achieve this, they must design new governance methods that provide the essential elements of stability and standardization, while ensuring spaces for experimentation, innovation, and growth of new markets.

Case-Studies

Methodology of the case-studies

The case studies in this paper explore the interactions between digital commons and infrastructure. They focus on novel forms of public support for infrastructure provision through digital commons. The case studies cover initiatives that specifically address an infrastructure-related challenge, i.e. support for the sustainable provision and maintenance of a service, as opposed to one-off research and innovation funding. The case studies have been guided by the following research questions:

• Under what conditions can Digital Commons (community-governed digital resources) participate in the provision of infrastructure (systems that have society-wide economic or social functions)? What type of infrastructural services are provided by Digital Commons?

• What is the nature of the relationship between the public and the private/civic spheres when infrastructure is provided by Digital Commons (procurement, partnership, funding, etc.)? Are there specific frictions or challenges? What does it mean for State support methods and modalities in general? What learnings can be drawn from this example? How do they differ from previous public intervention in the field?

• What does support as infrastructure mean for Digital Commons and their governance? How do Digital Commons adapt to public service obligations, sectoral regulations or accountability and transparency requirements?

In each case study, the context and actors involved are described. The study then identifies the infrastructure challenges and explains how the infrastructure is provided and what the governance structure is. It outlines the support from public actors and evaluates its sustainability. Finally, it links the findings to the research questions and discusses their implications.

Decidim (Nil Homedes Busquets)

Context

Decidim is a free and open, digital infrastructure for participatory democracy that is widely used around the world. More specifically, Decidim is a web environment (a framework) built in Ruby on Rails that allows anyone to create and configure a web platform to be used as a political network for democratic participation143. Decidim is currently used by more than 450 organizations in 30 different countries. Of these, 240 are cities and government organizations and 180 are social organizations. It has currently been implemented in various institutional contexts, including the City of Barcelona, the European Commission, the Government of Catalonia, the cities of Helsinki (Finland), New York (USA), the Brazilian Federal Government or the French National Assembly, among hundreds of others144.

Launched by the Barcelona City Council, the Decidim Barcelona project started in 2016 with the need to coordinate the Municipal Action Plan, a participatory project for the strategic planning of the next four years of the city projects. After being a success, two needs were raised: First, to have a more flexible software, that allows the city to have more functionalities to coordinate different participatory mechanisms, such as: being able to handle multiple participatory processes, assemblies, initiatives, consultations, etc. And secondly, the fact that other cities approached the City Hall and asked to reuse the software. Following this initial process, and due to the project's open and collaborative nature, the code was completely rewritten a year later. The decision was made not to develop from scratch but to reuse the code of the Consul platform (Decidim initially started as a fork of it), so that the platform could be reused by any other city thanks to a more modular architecture. Taking advantage of this modular architecture, external organizations have extensively extended Decidim, with more than 100 modules currently being developed by them145.

Thus, Decidim was initially created in response to a very specific need of the Barcelona City Council, and this, together with the lack of similar projects that met all the technical and political requirements that the City Council needed, are the reasons that led to the creation of Decidim. In Barcelona, it became the city's digital infrastructure for participation, and gradually its use spread around the world. Recently, Decidim was recognized by the Digital Public Good Alliance (DPGA) as a Digital Public Good.146 In this way, Decidim has become a public digital infrastructure. This infrastructure consists primarily of the participation platform itself, but also includes documentation, design elements, community modules and data sets. All these elements serve to compose a participatory democratic system within any organization147.

Governance model and community

The Decidim Association, created on February 16, 2019 in an extraordinary assembly148, following a participatory process149was conceived as the governance instrument of the Decidim community, that is, of the group of individuals and legal entities interested in the development, growth and improvement of the democratic infrastructure of digital participation based on Decidim. The Barcelona City Council has transferred the management and maintenance of the source code to the Decidim Association through an agreement signed between the Barcelona City Council, Localret and the Decidim Association. This public-common collaboration is an example of how to design the governance of digital commons. It is a unique example in the world of how to implement free software as a public policy with a community governance model.

The Decidim community is called Metadecidim and is made up of individuals and organizations that use the software and want to contribute to its improvement. Any person, company, university or public institution that has a relationship with the Decidim software is part of the community. In order to articulate the participation of all community members, the Metadecidim platform was established150. The Decidim project has always been developed in an open and collaborative way in all aspects. Anyone can participate in the process of democratically designing Decidim by simply accessing the “Propose new features”151participatory process in Metadecidim. Any member of the community can propose a new feature or suggest improvements to an existing one.

Support modalities and revenue model

The Decidim project has several sources of funding and resource generation, which are made public in detail in Decidim’s Sustainability Plan152. Today, the association is mainly funded by the Barcelona City Council and the Government of Catalonia, which make a transfer of resources to the association. These funds are invested in the dynamization of the community and the management of the contributions to the code.

In addition, every year the Barcelona City Council invests in new developments and improvements through public tenders carried out by companies of the Decidim ecosystem. Some of these improvements are customizations for the Barcelona installation153. However, there is also a part of this investment that goes to new functionalities and improvements for the main repository of the Decidim software. Thus, the city of Barcelona has financed most of Decidim's development so far.

Furthermore, a new partnership policy was introduced in 2022 to create a path for companies providing Decidim services to contribute to the digital commons that underpin their business model154. This policy requires companies to contribute 3% of their turnover from Decidim services. This mechanism relies on trust, as the association cannot enforce payment.

Conclusion

There are always tensions between the public, the private, and the commons. In the case of Decidim, the association represents the guardians of the commons and the community. It is the guardian of the original values and principles of the project. The private sector is represented by the companies that provide Decidim services, and the public sphere of the project is made up of all the governments that use Decidim and can contribute to its funding and governance.

The Barcelona City Council, as the main public partner of Decidim, has been a key factor in the success of the project, not only through its financial support, but also through its clear political support for the governance model of Decidim as a digital commons, recognizing the association as a central actor in the public-common governance model. This partnership includes regular meetings to align visions and define the roadmap between the association and the Council. The Government of Catalonia has also played a crucial role through funding and participation in governance. A key challenge for Decidim is to involve other global administrations in this governance model in order to establish Decidim as a universally shared commons and to diversify its funding. Relationships between digital commons and public institutions can also create tensions. For example, some of these tensions may arise from the current public procurement model, which, due to its rigidity, lacks the flexibility to adapt both to the unforeseen events typical of code development and to the needs of communities.

The diversification of Decidim's revenue model also relies on a greater involvement of the private sector in the maintenance and development of the software. Private sector involvement in Decidim presents a dichotomy: some companies exploit the commons for profit without contributing, while others engage constructively with the community, for example by becoming official partners or donating 3% of their revenue to the association. This "tragedy of the commons" scenario calls for mechanisms to encourage reciprocal contributions from private entities. An interesting development in the context of the digital commons is the possibility of new forms of cooperation between private sector organizations, sometimes even competitors, who can pool their resources through Decidim and sometimes even jointly bid for tenders to provide new services to the public. Such collaborations show the potential for private sector contributions to increase the robustness of the commons.

The lack of bureaucracy is pointed out as one of the limitations that have hindered the large-scale expansion of the digital commons155, in that sense a key challenge for Decidim's transition from a small community-based experiment at the local level to a global infrastructure for citizen participation is the need to further institutionalize some of the community's rules and processes, while maintaining the flexibility and openness of a bottom-up initiative. As the community expands, structured rules for participation and collaboration become increasingly necessary. Indeed, the growth of a digital commons, especially as partnerships with public and private institutions increase, requires the definition of stronger internal rules, for example to adapt to public requirements related to grants or tenders, but also to protect against potential extractive and fraudulent use of the shared resource. This includes defining a policy against the misuse of their tool by governments that do not genuinely intend to promote citizen participation.

The case study of Decidim provides key insights into how Digital Commons can effectively contribute to infrastructure provision. As a model of public-commons governance, Decidim is based on effective delegation to a community, represented by an association that ensures democratic participation, and the definition of rules to ensure the long-term sustainability of resource management. Several key lessons emerge from this case study. First, the long-term sustainability of Digital Commons requires ongoing funding to maintain the infrastructure, not just funding for new features. Second, fostering strong community and participatory governance is critical, and such engagement requires resources for collective deliberation processes. Third, the administrative adoption of digital commons by public institutions is crucial for scaling such projects and demonstrating political commitment. For those involved in the Decidim journey, their model can be replicated in various digital policies and services, promoting new forms of governance that transcend the traditional public-private divide. Political conviction and a commitment to community building are essential to the realization of these new governance models, which could play a critical role in addressing contemporary challenges, including the climate crisis, by putting democratic participation at the forefront of public planning and infrastructure development.

ZenDis (Nicholas Gates)

Context

Open Source Programme Offices, or OSPOs, create a strong enabling environment for supporting and deploying digital commons as part of society-wide infrastructures. By channeling and coordinating open source community efforts, and in turn creating channels for getting more funding and support back into the community, OSPOs can function not only as resource centers, but can scale support to bodies across entire jurisdictions, provide coordination with government policies and regulations, and ensure local innovation and contributions flow back to support the entire open-source ecosystem.

To better understand the potential of OSPOs for digital commons governance, this case study considers the case of Germany’s Zentrum für Digitale Souveränität (ZenDis) (‘Center for Digital Sovereignty’ in English). ZenDis is different from normal OSPOs in the way it goes beyond promotion of open source and policymaking, but also really focuses on actively working to bridge the gap between developer communities and government by helping to provide open source solutions for the public sector.

Challenge

While many OSPOs have had a lot of success in supporting and nurturing open source ecosystems,156 a challenge still remains in how you channel those efforts towards the specific needs of governments. OSPOs must still work within the confines of the strict legal requirements governments are subject to when procuring, implementing, and maintaining any software solution or solutions. This, in theory, creates more challenges for governments to support digital commons projects, where there is a presumption of more defined obligations around contribution, support, and governance of something which, in most cases, operates legally outside of government.157

One of the primary aspects of this challenge which ZenDis has encountered in working with the community and participating in community governance of open source projects – many of which ought to be regarded as digital commons – is the lack of guardrails.158 Because governments are subject to implementing standards defined by legislation, or upholding rules and regulations mandated by their own government or the EU, they need to be very intentional about what projects they support and who can participate in them. This is done with an eye towards mitigating risk and upholding security, while still being able to participate in and leverage the benefits of open source.159

Compliance is another issue. Most governments, including Germany, have strict requirements around which solutions you can use and what must be proven in order to adopt and maintain an IT project. While part of adopting open source in the public sector is changing some of these policies, either through rules or through legislation,160 change must also happen through innovations in governance. Therefore, ZenDis has been using their position in the open source ecosystem in Germany to try and figure out how to create more regulated spaces where they can still contribute to open source projects, but work with the community on their own terms. They do this in order to make sure that their contributions to, and the contributions they receive from, those projects uphold relevant rules and requirements for public procurement and maintenance.161

Description

The ecosystem ZenDis stewards provides a level of coordination that helps to ensure that there is healthy, community-based governance of open source projects being supported by the national government and adopted by local public administrations.162 Their office helps cement the role of key software and standards as digital commons while enabling a more active role for local and regional governments in the open source ecosystems present in their jurisdictions.163

Because of the aforementioned challenges, such as guardrails and compliance, ZenDis has sought to strengthen community governance models of open source digital commons projects, in order to enable public sector contribution while also adapting them to government requirements (e.g. procurement, cybersecurity) and facilitating rules-based access. Even beyond what other OSPOs have done, ZenDis has tried to build a thriving open source ecosystem and provide connections to public administrations. They also use those relationships to explore new models of governance that enable governments to contribute that mitigate risk.164

The most notable way ZenDis does this is through the platform Open CoDE, a joint platform of the public administrations in Germany which enables the exchange of open source code and software.165 As stated on the Open CoDE website: “The central repository of open source code is intended to promote reuse and collaborative work on public administration software solutions between administration, industry and society.”166 We argue that ZenDis, via the Open CoDE platform, provides policy infrastructure which is crucial for supporting and deploying digital commons as infrastructure.

Open CoDE is a platform built by ZenDis which is hosted as a mirror of GitLab. The goal of Open CoDE was to help create some of the guardrails mentioned above, as well as facilitate rules-based contribution to, and collaboration on, open source projects with dependencies in the public sector.167 At present, Open CoDE hosts more than 4,400 Users, 1,400 repositories, and 550 groups.168 Because of the need for a rules-based collaboration platform, it was stipulated from the outset when Open CoDE was launched in 2022 that you can only contribute through the platform if you are part of the public sector or have a partnership with a public sector entity. Projects can only be originated by the public sector, but the code is available and transparent to all since it is a mirror of GitLab, and contributions upstream from a downstream GitLab mirror still support project codebases outside of OpenCoDE.169

While Open CoDE is not the only platform public administrations and municipalities use, it has become the dominant alternative. As a regulated mirror, it helps create a solution for managing the tightrope of balancing community contribution and governance with the needs of public sector adoption and maintenance. In the view of the ZenDis team, it offers a reliable and necessary solution for facilitating the governance of many projects which can be considered digital commons in Germany. In particular, it has helped support the development of several other open source digital commons for the public sector, including openDesk.170

openDesk is a project to create an open workspace collaboration suite for use by German public administrations.171 With openDesk, the government is ostensibly trying to bundle many open source software solutions as part of a Phoenix Suite, which is being developed by a company called Dataport. The code has been hosted on Open CoDE and is being developed in the open, with collaboration from members of the open source community.172 Through Open CoDE, they are working to ensure that different companies can develop different components of openDesk and contribute to them, with the government playing a facilitating role in making sure they interoperate.173

Looking forward

Across all of their activities, ZenDis have been working to track the development of digital commons projects, with many examples emerging of projects with shared contribution from a number of regions and municipalities. Developing and launching Open CoDE has given public sector entities in Germany a forward-looking way to help support access and sharing rules for projects being used by the public sector, while enabling them to contribute upstream and participate in their adoption downstream.

Despite these successes, ZenDis faces some challenges in the road ahead as they seek to support digital commons. One challenge is around the issue of capacity. While they are collaborating with open source communities, the need still exists to improve technical talent in the public sector and enable more dynamic and active relationships with open source communities, particularly as new solutions emerge and are adopted by the public sector. They also need to do more work to involve civil society and strengthen the collaboration models for working with businesses in more defined ways.

Another issue is the issue of public funding and investment. As a result, they are actively identifying ways to act not just as a contributor but help the projects they support receive money, feedback, offers of support, etc. Moving forward, they want to build a closer relationship with the Sovereign Tech Fund,174 which funds a lot of critical open infrastructure in Germany and beyond. The goal of doing this is so that they can play a more strategic role in supporting and investing in highly open digital commons being used at scale as part of digital infrastructure all across Germany.

Conclusion

In summary, Germany’s ZenDis has taken crucial steps to shape the rules and norms of various open source software packages as digital commons through its Open CoDE platform,175 supporting the wide reach and impact of the public sector while also working with the communities that support the government. In ZenDis’ case, they also support coordination between open source developers and government bodies – including the Sovereign Tech Fund, which funds critical open infrastructure in Germany and beyond.176 This case study offers many lessons for how governments can facilitate participation in digital commons governance, even as they continually learn new lessons about how best to do so.

Despite some growing pains, the emerging evidence from ZenDis suggests a strong case for how the public sector can practically help support open source software as digital commons and procure digital commons as infrastructure, as well as make a case for open alternatives to proprietary solutions. By working with policymakers, OSPOs like ZenDis can play a key role in introducing digital commons as part of digital transformation strategy positions and blueprints for digital government transformation. This gives the government another policy lever to help standardize new business models for procuring and maintaining digital commons as infrastructure, helping to ensure the success of the model as a viable alternative to more restrictive commercial options.

DHIS2 (Lea Gimpel)

Many thanks to Scott Russpatrick and Mike Frost for sharing their insights and research on DHIS2.

Context

DHIS2 is a free, web-based, open-source platform designed for data collection, analysis, visualization, and sharing of aggregate and individual data. It is the leading health information management system, with implementations in over 80 low- and middle-income countries worldwide, covering 30% of the world’s population. It is also recognized as a digital public good in the DPG registry177. It was initially developed for the health sector but has been deployed in several other sectors, including managing education, agriculture, and logistics projects178. The history of DHIS2 spans almost 25 years: The web-based product was launched in 2006 and marked a shift from the standalone software DHIS1, which was first deployed in two South-African provinces in 1998179.

DHIS2 is a global software collaboration project managed by the University of Oslo’s Health Information Systems Programme (HISP). The HISP Center at the university serves as the central hub for coordinating software development, capacity building, and sharing knowledge and innovation within a global network that includes academic institutions, ministries of health, non-governmental organizations (NGOs), and funding agencies180. This international network comprises 17 in-country and regional hubs that offer continuous, direct support for implementation partners, usually Ministries of Health or NGOs.

Challenge

In low- and middle-income countries, DHIS2 fills a gap due to a lack of commercial providers for health information management services. Especially for last-mile services, there is no market for commercial providers, making public solutions necessary to give people access to fundamental services and ensure their right to health as enshrined in several human rights instruments, such as the Universal Declaration of Human Rights from 1948. Many low- and middle-income countries only have limited resources to develop, deploy, and maintain key software infrastructure. For this reason and because of their familiarity with DHIS2 as a software application deployed in some countries for decades, several countries have also started using DHIS2 in other domains (mentioned above). This development is aided by the HISP Network recognizing DHIS2’s potential as an innovation platform to provide social options rather than an application strictly focused on one application area. In the academic literature, social options are considered open, commons-based, generic resources that support social value creation for various issues181. Accordingly, in recent years, DHIS2 has been developed into an information management infrastructure that can underpin essential public service delivery.

Description

The software development, capacity building, and the sharing of knowledge and innovations in the global HISP network are managed by the HISP Center at the University of Oslo (in the platform literature, the HISP Center would take the role of a platform leader). As an innovation platform providing social options, DHIS2 faces a collective action dilemma: While broad participation of different stakeholders is essential in creating joint value for a community by enabling heterogeneous actors to contribute different assets to the collective interest, aligning diverging interests and activities requires orchestration and governance. Researched dynamics in platform ecosystems unveiled that open governance and deferring decisions regarding platform resources to contributors enable co-creation, spill-over effects, and innovations. However, governance structures that are too open can lead to a diversion of focus from the shared goals. Platform leaders - those who orchestrate contributions and steer the group toward a common goal - must simultaneously enable distributed autonomy and collective action.

That’s also the role of the University of Oslo’s HISP Center: It develops and evolves the governance mechanisms steering the global HISP network toward creating social value for and with the global network. HISP network hubs submit and vote on feature requests through an open roadmap process where everyone in the national and regional DHIS2 network participates. One key advantage of this governance structure is that the HISP Center can act as a mediator for individual goals of national and regional hubs, making sure that the product is developed in a generic, building-block manner that benefits everyone and makes in-country adaptation easy. National and regional hubs are encouraged to develop and share back their own modules and custom applications through open APIs, a software development kit, an App Hub and tutorials, thus capacitating local developers and providing a channel for decentralized innovation while also realizing local community benefits. One example is, for instance, the development and uptake of a Covid-19 tracker module, which was developed in Sri Lanka and subsequently implemented in more than 30 countries globally. All in all, the global community only contributes a fraction to the core product, but the HISP Center estimates that third-party contributors provide 15-20% of critical functionality in any implementation of DHIS2. Applications in the area of education and logistics are nearly exclusively dependent on third-party applications. This high rate of critical features developed outside of the core is also due to a strong community ethos, which is actively cultivated and results in the vast majority of users contributing back their generic innovations. The HISP Center aspires to increase the number of external contributions to the core by developing new contribution mechanisms.

Another benefit of this governance structure is its effect on trust-building. Since the University of Oslo maintains the product, DHIS2 is free of commercial interests and government meddling. This is a key component of the relationships the HISP Center built with numerous donors and supporters over the years.

Support modalities

Several donors, including NORAD, WHO, UNICEF, PEPFAR, etc., fund DHIS2. DHIS2 is only reliant on donor and research funding, without any other underlying business model. While this has been a successful strategy over many years, challenges persist: Donors tend to fund innovations and implementations, often neglecting maintenance of the core. Such funding is usually “catalytic,” with donors expecting to end their support at a certain point in time. However, infrastructure requires steady and reliable funding over many years without the perspective of meeting an “endpoint.” Third-party applications funded by short-term, catalytic funding or developed by the community without external support specifically face sustainability challenges, and DHIS2 currently explores commercial models to maintain these innovations indefinitely.

DHIS2 also struggles with the common free-rider problem of commercial software providers extending and building on top of their product without contributing anything back. In some cases, this leads to competitors offering the same services, such as data warehousing, coupled with additional features, effectively diverging donor funding from DHIS2 to commercial providers. Lastly, many Western countries' shifting political landscape and economic struggles have put international development funding under more scrutiny. It has already led to reduced budgets for development assistance in many countries, including Norway, Germany, the UK, and the Netherlands.

Conclusion

The journey of DHIS2 is a testament to the power of commons-based governance to work towards collective interests and create social value locally and globally. Such models seem especially pertinent in contexts in which market failures prevail or commercial models fail to address community values and public interest at large. The long history of DHIS2 shows the need to evolve governance structures over time to navigate the collective action dilemma and mitigate complexity in a rapidly changing technology landscape. In terms of relationship management with the HISP network and partners, capacity development and trust-building play an essential role. The emerging role of DHIS2 as a foundational infrastructure requires new funding vehicles and rethinking community governance mechanisms to address the free-rider problem, safeguarding the platform against commercial competitors and supporting non-commercial forms of value creation such as equitable healthcare for the many.

Scikit-learn (Cailean Osborne)

This case study stems from a research paper on the funding model of scikit-learn182.

Context

scikit-learn is a Python library that implements machine learning (ML) algorithms for classification, regression, and clustering, as well as related tools for data preprocessing, model fitting, model evaluation, and visualizations. Initially called scikits.learn, the project was started by David Courapeau as a Google Summer of Code project in 2007. After a dormant period, it was relaunched as scikit-learn in February 2010 by Fabian Pedregosa, Gaël Varoquaux, Alexandre Gramfort, and Vincent Michel, researchers at the French Institute for Research in Computer Science and Automation (INRIA). Nowadays, it is one of the most impactful and popular projects in the ML/AI OSS landscape, described as “the Swiss army knife of ML” due to its widespread use in research and industry.183 To date, the project has been maintained by core developers, mostly based at Inria, and a global community of volunteers.184 Since its first public release in 2010, scikit-learn has been supported by the Inria Foundation185 and a mixed funding model combining public research grants, corporate sponsorship, community donations, and as of November 2021 a €32 million grant announced in France’s artificial intelligence (AI) strategy.186

Challenge

At the time of scikit-learn's creation, there was a lack of open source libraries for ML. This gap in the ML ecosystem posed significant challenges for researchers and engineers, who sought to implement and compare various ML algorithms without the burden of writing code from scratch or relying on scattered, inconsistent, and often poorly maintained implementations.187 Nowadays, OSS is indispensable to ML/AI research and innovation,188,189 and scikit-learn is used by millions of researchers and engineers across the world for building predictive models for diverse applications, from research fields like bioinformatics and climate science to industry use cases such as fraud detection and stock price prediction. However, the maintainers explained that its sustainability and impact were not inevitable, and the project faced a number of challenges, from the under-development of the scientific Python ecosystem in the early 2010s to pressures stemming from industry dominance in AI research and development190. As explained below, its community-approach has been crucial in ensuring that scikit-learn remains a state-of-the-art digital public good for ML researchers and engineers that is available for free under its permissive BSD license and has remained independent of the strategic goals of a single vendor.

Description

scikit-learn is provided for free as a Python library, available for anyone to use, modify, and distribute under the permissive BSD license. The project is maintained and developed by a core team of developers, primarily based at Inria, and a global community of contributors who voluntarily dedicate their time and expertise to improve the library.191 The community contains volunteers from across the world and companies192. The majority of contributions come from researchers and developers in the USA, India, Germany, France, the UK, Canada, China, Japan, Switzerland, and the Netherlands. The maintainers lay great importance on this diversity and make efforts to cultivate the project’s identity as the collective effort of a global community. For example, Julien Jerphanion, a maintainer, explained that, “The major part of the labor is based on benevolence of people working in their free time and not asking to get paid.” This community-based model ensures that scikit-learn remains responsive to the needs of its diverse user base, which includes researchers, data scientists, and software developers, amongst others. The community-driven approach also aligns with open science principles, which emphasize transparency, collaboration, and knowledge sharing. According to Gaël Varoquaux, a co-founder, the project seeks to provide a public alternative to the tools offered by profit-oriented industry giants and while the project is funded by public and private entities, decision-making about the project will and must always “come from the community.”

Support modalities

scikit-learn has benefitted from several sources of funding, including public research grants, commercial sponsorship, micro-donations, and a €32M grant announced in France’s AI strategy.193 Each type of funding has presented unique advantages and challenges, and have funded different aspects of the project under different timelines. Since 2010, according to François Goupil, the community manager, INRIA has provided around €1.5 million in support via staff salaries, public research grants, office space, computing resources, and event sponsorship. As Olivier Grisel, a maintainer, noted, “[Public funding] is not new with the AI strategy.” The project also benefits from micro-donations from the community via NUMFOCUS and student projects have been sponsored by the Google Summer of Code programme. Adrin Jalali, a maintainer, explained that while “the major stuff is not funded through NUMFOCUS”, it has been useful in funding marketing, events, and so far one internship for underrepresented groups.

In 2018, the scikit-learn consortium was established under the Inria Foundation to stabilize funding, to secure employment for maintainers, and to develop new features.194 Companies join via annual memberships in three tiers: silver (€30,000), gold (€50,000), and platinum (€100,000). The Technical Committee elaborates a strategic technical roadmap for the project, including gathering feedback from the community; while the Advisory Committee advises on various topics, such as the consortium's membership and financial status. To date, it has included Dataiku, Microsoft, Nvidia, Intel, AXA, Boston Consulting Group, BNP Paribas Cardif, Hugging Face, Fujitsu, and Chanel. Other companies, such as Quansight Labs, have indirectly funded the project by sponsoring maintainers. Through their sponsorship, companies gain a voice in scikit-learn's development via the Technical Committee and Advisory Board. While the maintainers appreciate the industry use cases and needs that their consortium members share, some sponsors have sought to influence the project's direction or enhance their reputation in ways that conflicted with project's norms. To manage these demands, the maintainers employ governance protocols that limit funders' influence and prioritize community control. The Technical Committee cannot override decisions with rough consensus amongst maintainers. Adrin Jalali, a maintainer, explained, "If there's two thirds majority in cast vote, for whatever vote we do, then the technical committee doesn't step in…because the power is given to the [maintainers] and that's by design". This example highlights the importance of governance protocols to safeguard the community interests whilst being funded by companies with private interests.

In November 2021, France's AI strategy, "Stratégie IA", announced a €32 million grant for the duration of five years to support the development and maintenance of the scikit-learn project; adjacent Python libraries for data science developed by the maintainers, such as skrub, joblib, and fairlearn; and the development of an open-source platform for data science. Through this funding, the French government aims to enhance French competitiveness in AI R&D, facilitate AI adoption throughout the economy,, and support the digital sovereignty of France and Europe. The maintainers expressed their gratitude for the government's financial support and recognition. Adrin Jalali, a maintainer, emphasized the stability the funding brought to the project, allowing them to make long-term plans for recruitment and the technical roadmap. François Goupil, the community manager, especially praised a requirement to acquire matching funds from both public and private sources throughout the EU: “I think it would be dangerous for us to be exclusively funded by the private sector or to be exclusively funded by the French government, because we have many good contributors who are not French.” However, challenges emerged in aligning policy goals with the project's core strengths and community norms. For example, disagreements arose around expanding scikit-learn into deep learning, developing OSS tools without competing with domestic companies, and the slow pace of fund disbursement which could be critical for more financially precarious OSS projects. Some consortium members raised concerns about the potential politicization of the project, with French policy goals like digital sovereignty probably not being well received by the global community. Despite these challenges, the maintainers remained committed to preserving the community ethos of scikit-learn and commended their constructive interactions with the government officials. In particular, the government's multi-stakeholder process involving experts from research and industry was useful for refining the grant, demonstrating the value of multi-stakeholder consultations for designing grants that balance policy goals with the expertise and needs of OSS developers.

Conclusion

This case study highlights the role of diversified funding, including public research grants, community micro-donations, and private sponsorship, in sustaining community-governed OSS projects like scikit-learn. It sheds light on the respective benefits and challenges of being funded by public institutions like the French government and private companies. On the one hand, commercial sponsors understand the importance of maintenance and expose the maintainers to industry use cases and challenges, but in some cases commercial sponsors try to influence the project in ways that are perceived to be at odds with the community norms. On the other hand, the government is providing long-term support for the project in the public interest, but the slow pace of the French government and the initial neglect of funding maintenance were challenges that the maintainers had to address. The community-based governance of scikit-learn has been critical to ensuring that the project remains focused on serving the needs of its users and contributors, rather than being driven by commercial or political interests. While it is still too early to evaluate the impact of the French government’s €32 million grant, the political will of the French government to support scikit-learn and the digital commons for data science through such a substantial, long-term grant, as well as to adapt the funding package based on dialogue with the scikit-learn developer community, is commendable and should be noted as a potential blueprint for future funding interventions by other governments.

Overall, this case study contributes to the understanding of infrastructure provision through digital commons and the role of public actors in supporting these initiatives, providing insights into the design and implementation of a public-private OSS funding model, the governance mechanisms necessary to safeguard community interests, and the potential for such models to be replicated in other digital commons projects. The success of scikit-learn's funding and governance model serves as a template for other community-led OSS projects seeking long-term sustainability and focus on their users and contributors.

European Open Science Cloud (Roksana Wilk)

Context

The European Open Science Cloud (EOSC) aims to create a federated and open multi-disciplinary environment for European researchers, innovators, companies, and citizens to publish, find, and reuse data, tools, and services for research, innovation, and education. It promotes seamless access and FAIR (Findability, Accessibility, Interoperability, and Reusability) management of research data and digital objects. The genesis of EOSC can be traced back to the European Commission’s 2015 communication, "A Digital Single Market Strategy for Europe,"195 which laid the groundwork for integrating Europe's digital research infrastructure. The initiative gained momentum with the 2016 communication, "European Cloud Initiative – Building a competitive data and knowledge economy in Europe," where the European Commission proposed the creation of the European Open Science Cloud196. In 2017, the EOSC Declaration outlined guiding principles and invited stakeholders to endorse and commit to contributing to the EOSC’s development197. The implementation phase officially began with the establishment of the EOSC Governance Board and Executive Board in 2018. The EOSC Strategic Implementation Plan198 and the EOSC Partnership under Horizon Europe199 are crucial milestones, guiding the initiative from 2021 onwards.

As an initiative set up to facilitate a practical implementation of open science and overall support the overall scientific process in Europe and Beyond, it gathers a significant number of important actors and stakeholders:

• The European Commission: As the driving force behind EOSC, the European Commission provides funding, policy guidelines, and coordination. Key roles are played by the Directorate-General for Research and Innovation (DG RTD) and the Directorate-General for Communications Networks, Content, and Technology (DG CONNECT).

• The EOSC Association: Established in 2020, the EOSC Association includes research-performing organizations, service providers, and industry representatives. It oversees the strategic implementation and operation of the EOSC ecosystem, ensuring alignment with the European research community’s needs.

• National Governments and Funding Agencies: National governments and funding bodies like the French National Research Agency (ANR) and the German Research Foundation (DFG) offer financial support and policy alignment. They facilitate infrastructure development at the national level.

• Research and Education Networks (NRENs): Organizations like EGI, EUDAT, OpenAIRE, GÉANT and their technology partners provide the connectivity backbone of EOSC, linking institutions across Europe and enabling seamless data sharing and access to services.

• Research Infrastructures: Institutions such as CERN, EMBL-EBI, and ESA contribute domain-specific expertise, large-scale datasets, and advanced research tools, ensuring EOSC aligns with cutting-edge scientific needs.

• Libraries and Data Repositories: Entities like LIBER and university libraries advocate for open access and ensure data repositories comply with FAIR principles. They also support and train researchers to engage effectively with EOSC.

• Academic and Research Institutions: Universities and research institutions actively contribute data, develop tools, and adopt EOSC infrastructure to enhance research efficiency. They engage in pilot projects and community-building activities.

• Research Communities and User Groups: Scientific communities, including life sciences (ELIXIR), environmental sciences (ENVRIV, Copernicus), social sciences (CESSDA and many others, drive requirements and service development. Their feedback ensures EOSC remains relevant and valuable to users.

Challenge

Recognized by the Council of the European Union as part of the European Research Area's 2022-2024 policy agenda200, the EOSC seeks to deepen open science practices and integrate with other sectoral data spaces in the European data strategy. Its deployment is expected to enhance research productivity, innovation, and trust in science.

The European Open Science Cloud (EOSC) exhibits infrastructure characteristics notably by setting norms and standards for data circulation. The development of this infrastructure involves creating shared technological backbone infrastructure offering data, tools and services for the scientists and science-facilitators, while addressing European open science policy aspects and the development of sustainable business models, and supporting the adoption of best practices. The objective is to facilitate the exchange of capacities, such as data storage and computing resources, across disciplines and countries, fostering a more integrated and efficient scientific community. By developing pilots that integrate services and infrastructures, EOSC supports interoperability across various scientific domains. Engaging a broad range of stakeholders, it builds trust and skills essential for adopting open scientific research practices. EOSC reduces fragmentation and improves interoperability between existing data infrastructures, enabling sharing and reuse of complex data across different domains and formats. This approach leverages existing resources to create a reliable open-data environment, ensuring that data from publicly funded research is open and that incentives for sharing data are created.

Description

For the time being, community-based governance is adapted for the EOSC instead of classical state or private sector provision due to the need for a diverse and inclusive approach. Initially, the initiative was fragmented and led by different EU Commission departments, but has since evolved into a more coordinated effort based on the tri-partire governance structure. At the same time, entities building and leveraging the EOSC activity are transforming into the “EOSC Nodes Federation”. Technicalities and practical implementation details for that model are under development guided by the EOSC Association with the support and resources of other stakeholders constituting the EOSC initiative.

These multi-stakeholder models include various actors, from the European Commission to the research community. In both of them the bottom-up approach ensures that initiatives are driven by those directly involved in implementation, aligning diverse national and sectoral goals. However, there is an ongoing discussion, whether some of the organizational and governmental topics like ‘possible money transfer mechanism and supported business models in EOSC Nodes Federation’ should not be decided using the top-down approach with a considerable amount of community consultations and discussions.

Support modalities

The implementation of the European Open Science Cloud (EOSC) began in 2015 with efforts by the European Commission to align and coordinate with various stakeholders in the European research landscape. In its initial phase (2018-2020), the European Commission invested around €250 million through Horizon 2020 to prototype EOSC components and set up interim governance to prepare for post-2020 strategies. The current phase (2021-2030) is guided by a Strategic Research and Innovation Agenda (SRIA), developed with the EOSC community, and is focused on a stakeholder-driven approach.

The project-based funding of the EOSC requires researchers to form consortia to apply for specific calls. While this funding approach fosters innovation by promoting competitive grants and cross-disciplinary partnerships, it poses sustainability challenges, as the infrastructure development is tied to specific projects with limited timelines. However, stable activities are maintained through statutory funding from public institutions, with services like software and user support provided on a best-effort basis. Additional support comes from organizations like CERN and NASA, as well as national funds dedicated to digitalization and cybersecurity, which can align with EOSC objectives.

A co-investment of at least €1 billion from EU and non-EU partners is planned for the next seven years. Project officers from the European Commission collaborate with research communities to identify key areas for future funding, while the progress is overseen by the tripartite governance structure.

Conclusion

The relationship between public, private and civic spheres within the EOSC is characterized by a dynamic interplay where different research communities, organized as consortia, collaborate to provide various components of the EOSC infrastructure. This collaboration is underpinned by public funding primarily through European Commission initiatives, such as Horizon 2020, alongside national government contributions. Policy instruments used in the case of the EOSC differ significantly from previous public interventions in the field. Instead of a top-down approach, the EOSC was designed to employ a bottom-up, community-based governance structure that prioritizes inclusivity and collaboration with research institutions.

However, the EOSC faces several challenges. The reliance on project-based funding introduces sustainability issues, as infrastructure development is often tied to specific, time-bound projects. This can lead to fragmentation and a lack of continuity in services and support. Additionally, the multi-stakeholder governance model, while inclusive, can lead to complexities in decision-making and coordination.

This case study reveals important insights for public support methods and modalities. It underscores the necessity of a mixed funding approach, combining competitive grants with stable, long-term funding commitments to ensure sustainability. Moreover, it highlights the value of a stakeholder-driven governance model that integrates diverse perspectives and expertise, thereby enhancing the relevance of the infrastructure for the public. An increasingly decentralized approach is envisioned for the future of the EOSC, as the EOSC Nodes Federation will allow more flexibility in terms of financial mechanisms and business models within the EOSC.

Final conclusion

The first part of this publication has shown that current public value maximization strategies for public digital infrastructure adopted by policymakers emphasize the role of the state as an "entrepreneurial state" that promotes a digital environment based on openness and generative interoperability. This approach prioritizes the creation of digital infrastructures that support multiple public functions and enable multiple derivative uses, recognizing the importance of openness and non-exclusivity. Policymakers propose various forms of public ownership to ensure the public character and societal benefits of these infrastructures, all of which include some digital commons as a mechanism for infrastructure provision.

The case studies analyzed in this paper confirm this nuanced and multifaceted approach to public digital infrastructure, using digital commons and various forms of public ownership to maximize societal benefits and ensure inclusive, open, and interoperable digital ecosystems. The case studies of ZenDis and sickit-learn allow to better illustrate the ecosystem strategy followed by public institutions aiming to strengthen the stack of OSS increasingly connecting information systems and devices. The case of Decidim, created by Barcelona as a new type of commons-based institution, shows what kind of partnerships can be built to provide alternative spaces for online public speech and debate. DHIS2 - a software component used by many governments to develop their health information management systems - illustrates how public digital infrastructure can be collectively managed and produced, allowing mutualization of resources and capacity. Finally, the European Open Science Cloud is a unique example of a pan-European attempt to build a shared infrastructure with a community of these researchers, not only redefining the tools and processes used by researchers, but also developing an infrastructure embedded in the values and principles of open science.

The case studies provide various insights into the necessary conditions for digital commons to provide infrastructure, especially in terms of long-term funding, which as shown by the review of literature in this paper, communities struggle to provide. It also shows the diversity of relationships that can be built between public institutions and digital commons, with consequences on the governance of digital commons.

Necessary conditions

The case studies show that the potential of Digital Commons to contribute to public digital infrastructure provision hinges on several key conditions. Firstly, there must be a clear need that aligns with a public function, which implies that the Digital Commons should contribute to achieving public goals. These goals could be related to, for example, a Sustainable Development Goal or the enjoyment of a fundamental right. Secondly, while the degree of formalization of the commons can vary, some level of it seems necessary and inevitable for a successful public–commons partnership. This formalization could take various forms. For instance, it could be achieved by setting up an association or by integrating the Digital Commons into a public research institution. This formal structure provides a framework for collaboration with the (other) public institutions. Thirdly, the long-term sustainability of Digital Commons is contingent on continuous funding. This funding is crucial not just for the development of new features, but also for maintaining the existing infrastructure and facilitating collective deliberation processes. Finally, for the Digital Commons to scale as a PDI political will and favorable political climate are indispensable. Without the backing of the political establishment, the Digital Commons may struggle to achieve the necessary scale and impact.

Nature of the relationship

When considering the relationship between public institutions and the digital commons, it is crucial to recognize that public institutions are not a homogeneous category. Different institutions can assume different roles and perform different functions in relation to the commons. Nevertheless, the relationship between public institutions and the digital commons in the context of infrastructure provision revolves primarily around governance and funding. In terms of governance, there are different setups. In some cases, governance is fully delegated to a community, as seen with Decidim, where state institutions play a minimal role in governance structures. Alternatively, there can be co-governance, such as with the European Open Science Cloud (EOSC), where tripartite governance involves strategic coordination among the European Commission, the EOSC Steering Board representing participating countries, and the EOSC Association representing the research community. In other instances, the public institution (e.g. a university) remains the central governance hub, as is the case, for example, with DHIS2. Funding and other forms of support are another critical aspect of the relationship between the public institutions and the commons. While donations and grants are popular, more public funding is typically available for innovation rather than maintenance of the commons. Public funding for digital infrastructure in the form of projects often poses sustainability challenges. Public tenders can be used for new developments and improvements of the commons, but these come with challenges in adapting to public procurement rules and the risk of preferential treatment or “picking winners”.

Consequences for the digital commons

When digital commons receive support as infrastructure, it brings about significant implications for their governance and operation. One of the primary concerns, especially in digital commons projects initiated by the community, is the risk of ‘statization’ and politicization. This fear stems from the potential for increased state control or political influence that could alter the original objectives or operations of the commons. When digital commons assume the role of PDI it impacts their relationships with both the state and market actors. This could manifest in various ways, such as dealing with private funding or navigating attempts by private actors to influence the project agenda. The transition to a PDI role may also introduce new requirements and expectations. These could pertain to transparency, governance, or technical requirements, necessitating adjustments in the commons’ operations and strategies. Scaling up presents its own set of challenges. Governance, in particular, can become a collective action dilemma. While broad participation of different stakeholders is essential to create common value for a community, aligning divergent interests and activities requires careful orchestration and governance. Finally, there is the perennial challenge of “free riding” by other actors. This is a common issue for digital commons, where commercial software vendors may extend and build on the product without contributing anything in return, leading to competition and potential dilution of the commons’ value.