Governing as Commons or as Global Public Goods: Two Tales of Power

Commons and global public goods have become popular concepts in academic debates on governance. Moreover, these concepts are no longer the exclusive domain of economists. Different disciplines (such as legal and political theory) have appropriated them in their own specific ways. The result of this popularity, however, is that they are often confused or used in ways that muddle their distinct characteristics. In this article we propose some distinctions to clarify the use of these concepts. First, we will show how what were initially social scientific concepts started being used in a more normative way. Second, we will subject the writings of Elinor Ostrom and Inge Kaul and colleagues to a discourse analysis. This means that we will show that some normative assumptions were already present in the concepts of ‘commons’ and ‘global public goods’. We take it that, although Ostrom and Kaul are often read as social scientists, it is both possible and fruitful to read them as proposing two very different visions of power in a globalizing world. In a third section we then demonstrate more concretely what these visions look like. Finally, we conclude by looking at the possible advantages and downsides to both models.


INTRODUCTION
issues of effective provision (how the benefits are generated). When an identified problem is framed as the result of a lack in the production of GPGs, the conventional response consists of states and international institutions working together in estimating the demand and determining a division of production, focusing on one specific good.
Commons (in the sense of areas containing CPRs) would thus be in need of a different governance regime than GPGs. But framing problems as GPGs has occurred in many instances where a commons framework would have been equally appropriate (Bryner 2015, 41). As we will explain below, both commons and public good theory have integrated normative elements.
We, therefore, argue that framing issues either in terms of commons or in terms of GPGs, should no longer be seen as based solely on traditional, objective criteria of excludability and rivalry, but that it has important normative components as well.

Normative discourses
A single property or area can produce several kinds of goods; a forest can produce wood for private contractors but also provide room for recreation. We must also consider that resources can be held under several types of governance structures: for example, the aforementioned forest could be privately or publicly owned (Acheson 2011) and by local or international institutions. Considering something as either a commons or a GPG entails more than a simple analytical exercise. Even though the results of overexploitation and underprovision are comparable in as far as they both lead to a lower than optimal presence of the good in question, the perceived origin of the problem and the possible solutions are radically different. In this respect, the definition of a global issue as either a commons or a GPGs problem, determines to a great extent the institutional arrangements, and governance mechanisms applied to tackle it (Young 1989;. Whether a good is overexploited or underprovided is a seemingly objective question, but it requires in fact a normative judgement: does the food crisis in the Global South endure because of a lack of international aid, or are the farmlands in developing countries overexploited or cultivated in the wrong way? Do we witness outbreaks of contagious diseases because of overpopulation or because of poor housing conditions? Is the climate changing because forests are destroyed or because our industries produce excessive amounts of CO2? Such framing shapes the interpretations of issues and events (Matthes 2012) and, thus, also the possible solutions for these problems.
Although both the commons and the GPG frameworks can still serve a useful descriptive and analytical function, they are also actively used to promote certain solutions for perceived problems in a normative sense. Unsurprisingly, these concepts have gradually come to be regarded as prescriptive, going beyond the description of what is or could be, to sketch what ought to be in the international arena. As Menatti (2017) notes, 'commons have been often claimed as a new institutional and organisational framework -both local and global -in defining a non-capitalist society'. Commons have become a rallying call for a collection of social movements such as Occupy Wall Street, the Spanish Indignados (Mouffe 2013) and the Rodotà Commission (Rodotà 2013). What these movements have in common is that they all argue for increased self-governance and participation by citizens (or 'commoners' as they are Organization (Woodward et al 2002), and the European Union (European Commission 2002) all branded the provision of GPGs as a new policy challenge of its own. This new challenge would require closer and more comprehensive international and perhaps even supranational cooperation, with increased abilities for international institutions to ensure that all relevant states and actors contribute.

POWER
Now that it is clear that these concepts are being used in a more normative way, we will show what normative assumptions were already present in the original concepts. More precisely, we will argue that Ostrom and Kaul propose two different models of power in a globalizing world. And if we want to come to grips with Ostrom and Kaul's hidden assumptions concerning power, we would do well to start with their views on the nature of power.
What is immediately apparent is that the literatures on GPG and commons reveal very different conceptions of the nature of power. Inge Kaul thinks that Ostrom is insufficiently aware of the different forces that operate in the global sphere (Kaul et al. 2016). Consequently, proponents of GPGs criticise commoners for downplaying questions of force, coercion and violence. Defenders of the commons often return the compliment, noting that proposals for organising the provision of GPGs rely on top-down enforcement, as opposed to bottom-up cooperation and communication (e.g. Quilligan 2012). Promoters of GPGs mistakenly assume that a degree of force, coercion and manipulation are a necessity, while in fact people are perfectly capable of working together and hence achieve better results that fit local circumstances (Ostrom 1992).
These stances neatly overlap with two influential positions on the nature of power in political theory. On the one hand, thinkers like Thomas Hobbes and Max Weber conceive of power as power-over. In the words of the latter: power is 'the probability that one actor within a social relationship will be in a position to carry out his own will despite resistance, regardless of the basis on which this probability rests' (Weber 1968, 53). Consequently, violence, coercion and manipulation are treated as modalities of political and social action. Moreover, as states have a monopoly on violence and have most of the means to carry out their 'will', they can be seen as the embodiment of power-over.
On the other hand, there is the view, famously defended by Hannah Arendt, which involves an understanding of power as power-with. This form of power comes into being only when people act in concert and when they engage in unconstrained communication (Habermas 1977, 3). Here, power is not something that a certain actor 'has' and exercises over another, but exists 'in between' people. In other words, power is relational as opposed to unilateral (Penta 1996, 212). This entails that violence, coercion and manipulation appear as the opposite of power. As Arendt argues, 'power cannot be stored up and kept in reserve for emergencies, like the instrument of violence, but exists only in actualization', that is, 'where word and deed have not If we look at the writings of Elinor Ostrom, it quickly becomes apparent that she adheres to the notion of 'power-with'. Her research on CPRs, she argues, challenges 'the Hobbesian conclusion that the constitution of order is only possible by creating sovereigns who then must govern by being above subjects, monitoring them, and by imposing sanctions on all who would otherwise not comply' (Ostrom, Walker and Gardner 1992, 414). In other words, people can cooperate and keep credible commitments without there being an external enforcer. However, there are some important conditions for a successful constitution of order: first, people have to be able to communicate and preferably do so regularly and face-to-face; second, people should able to freely choose an 'internal sword' -institutional stability thrives when participants are able to establish common norms and agree on sanctioning mechanisms, in order to punish those diverging from these norms (Idem, 414); and third, it is vital that 'most of the individuals affected by a resource regime can participate in making and modifying rules' (Ostrom 2000, 150). To condense the argument: in these arrangements power does not rest in an external institution or a centralised authority with the ability to impose and enforce rules of action (Ostrom 2012), but exists in between people in a horizontal contract they conclude among themselves and keep alive through communication and their shared action. 2 If we turn to Inge Kaul's work on GPG, we find a different conception of power. Similar to Ostrom, she acknowledges the important task that institutions perform in the creation of order.
However, the way Kaul conceives these institutions is somewhat different. Institutions and regimes, she argues, can be considered as intermediate public goods, as they contribute towards the provision of final public goods. One indispensable task that they perform is that of monitoring and surveillance, since the main problem affecting the provision of public goods is that of free riding. This implies that in the absence of institutions directing or constraining these individuals (or, on the global level, states and corporate bodies), it is impossible to ensure cooperation. In short, it seems that the steering role of the state, its ability to discipline free riders either through norm setting or through the exercise of coercive power, is of essential importance to the provision of public goods (Kaul, Grunberg and Stern 1999b, 6-13). However, on the global level, state failure is a given 'due to the absence of a global sovereign' (Idem,15). Therefore, the nature of power changes as well. As Kaul writes, 'conventional hard power may have to be combined with soft power to yield a 'smart power' strategy that enables states to achieve a successful policy outcome' (Kaul In Advance, 23). However, this does not necessarily mean that Kaul veers toward the Ostrom position. First, the exercise of political pressure and coercion by states still has an important role to play in her scheme for the production of GPGs (Kaul 2013, 11). Second, power that is exercised non-violently and noncoercively can still constitute a form of 'power-over'. This happens when communication and participation are only admitted to the extent that they help actors achieve their goals. It persists where communication is not valued in itself, but only as a means for attaining a specific objective (Habermas 1977, 4). This is, in our view, the case in GPG discourse: communication and participation do play a role, to be sure, but only to the extent that they help in the attainment of a certain goal -namely, the provision of GPGs (Nordhaus 2005, 93). In this sense there is some truth to Joel Wainwright and Geoff Mann's claim that GPG discourse manifests a wish for a Climate Leviathan. This is 'a regulatory authority armed with popular legitimacy, a panopticon-like capacity to monitor and discipline carbon production, consumption, and exchange, and binding technical authority on scientific issues' (Wainwright and Mann 2012, 6).

THE ROLE OF THE STATE
From an institutional perspective, both commons and GPG theories reflect a deep frustration with the dominant state-centric system of governance. In both streams of literature, the opposition seems to be with the state understood in its more narrow organizational and bureaucratic sense. The state is seen as an obstacle. Both models indeed contrast with the Westphalian model, where the ultimate power (sovereignty) is situated in a state ideally organised in a hierarchical and centralised manner, and exercising control over most -if not all -policy-areas in a territorially defined jurisdiction. Both commons and GPG models recast this traditional (if not mythical 3 ) idea of the Leviathan state as a basic regulatory power with the monopoly on the legitimate use of force within well-defined territorial boundaries (Weber 1978, 56), albeit in very different ways. In this section, the 'state' represents the abstract notion of public authority that was and is still deemed necessary to coerce the behaviour of rational individuals and solve classic social dilemmas -be it of overexploitation of CPRs or underprovision of GPGs. It is as a reaction to this traditional and hegemonic model of governance that the commons and GPGs have re-appeared as counter-narratives.
On the one hand, commons scholars Ostrom 2009;Weston and Bollier 2013) challenge the centrality of the state because they reject the pessimistic idea that communities are incapable of managing shared resources collectively, without control and supervision by public institutions. Since Garett Hardin's 'Tragedy of the Commons' (Hardin 1968), it was indeed assumed that CPRs would be completely depleted unless enclosed through private property, or, failing that, public regulation. Commons scholars and activists, in contrast, argue that there is a third solution, which is neither 'all-private', nor 'all-public': the selfinstituting and self-controlling practice of communities themselves. Elinor Ostrom was one of the first authors to forcefully and successfully defend this argument in Governing the Commons (Ostrom 1990). Her acceptance speech for the award of the Nobel Prize (Ostrom 2009 Commons represent in that sense a new 'third way' beyond the dominant market-state dilemma. However, this is not to say that commons scholars aim to completely eschew the state or constitute a substitute for the state. While there are diverging ideas and political perspectives in the debate about the role of the state in the governance of commons, the consensus is that most commons will need to be formally recognised as autonomous institutions by public authorities in order to survive. In her book, Governing the Commons, Ostrom puts forward as a seventh design principle that '[t]he rights of appropriators to devise their own institutions [should] not be challenged by external governmental authorities' (Ostrom 1999, 101). Indeed, 'if external governmental officials presume that only they have the authority to set the rules, then it will be very difficult for local appropriators to sustain a rule-governed CPR over the long run' (Ostrom 1999). As Jane Mansbridge showed, this is not an 'anti-state message' (Mansbridge 2014, 8).
Although Ostrom clearly stands against a state-centric tradition, external government authorities can in fact be instrumental in 'nesting' local commons within higher levels of decision-making. This is in line with Ostrom's concept of polycentrism (as opposed to 'monocentrism') where each centre of decision making -at the local, regional, state and global levels -is formally independent of each other but enters into cooperative undertakings (see below). Commons activists today also recognise that the regulatory state continues to be 'the On the other hand, the idea of public goods was originally developed in the field of Keynesian economics to provide legitimacy to a greater involvement of the state (Samuelson 1954). In the project defended by Inge Kaul and her colleagues at UNDP, '[n]ation states form important core elements of the international community' (Kaul 1999b, 15). Even if they concede that states alone are no longer sufficient to deliver public goods that transcend national boundaries, they argue that these remain the dominant actors. Indeed, state intervention is still considered as indispensable in the financing and provision of public goods at the national level.
However, supranational and top-down mechanisms appear necessary to coerce free-riding states into the provision of much needed GPGs. Specifically, GPG supporters point to the absence of a 'global sovereign' or a state-like entity capable of enforcing contribution of GPGs by all states.
That is the crux of the matter. This is why they call for more international cooperation between states, for example through international organisations such as the UN. They also warn that the principles of state sovereignty and state consent represent obstacles for the effective provision of GPGs -the 'basic problem [that] underlies all others' (Zedillo and Thiam 2006, 3). Much like the individuals in the prisoners' dilemma, it is expected that sovereign states acting in total independence will defect from cooperation, unless coercive mechanisms are introduced (Kaul 1999b, 8). What this analogy highlights, is that, in order to coerce free-rider states, we need to transpose certain domestic strategies to the global level (Constantin 2002, 81). For some legal scholars, for example, it seems the solution is to impose inter-state cooperation by 'design [ing] punishments that are sufficient to induce compliance' (Trachtman 2012, 161). What is striking in comparison to the bottom-up commons framework, is that GPGs call for closing the jurisdictional gap at the global level with more inter-state cooperation from the top. As Martin Deleixhe concludes, the organization mode of GPGs 'is closer to an inchoate world-state than to a polycentric federalism' (Deleixhe 2018).

THE SCALE OF GOVERNANCE
Both models thus present a framework in which the role of the state is much reduced and different. They suggest governance should take place differently and at different levels, but provide different answers to questions of what these levels should be and what the appropriate 'scale' of these alternative governance systems should be. 'Scale' has generally been used in an analytical sense (Gibson, Ostrom and Ahn 2000, 218). 4 However, scale can also be used in a more normative sense, referring to two elements: (a) the question of the most appropriate level(s) (or spatial construction more generally) for governing certain goods or issues; 5 and (b) the hierarchy among the different levels in decision-making. It is the responses of the commons and GPG governance models to these questions that we study here. As critical geography has demonstrated, the answers to such questions are 'socially constructed' and 'dynamically active', rather than objectively determined and fixed (Rudestam, Langridge and Brown 2015). 6 We limit ourselves to the ideal spatial construction and relation among levels of governance set out by these models, excluding a discussion of, for example, their conceptions regarding the appropriate temporal scale. 7

Appropriate Level of Governance
Commons theory has its roots in the study of local communities collectively managing a CPR, such as a pasture or an irrigation system. These roots have two important implications in determining the most appropriate level for governance. Firstly, since its origins the theory has established a close link between the governance system and the resource system it is called to govern. Giordano aptly states that '[a]t its most fundamental level, the problem of the commons revolves around humans, their environment, and the spatial relations between the two' (Giordano 2003, 365). Therefore, as resources are increasingly viewed as part of larger ecosystems, commons theory has increasingly reflected the need for multi-layered, cross-scale governance systems. Secondly, commons theory has attached much importance to trust and reputation, as variables altering the traditional economic models which suggested cooperation between self-interested, rational individuals was impossible (Ostrom 1998). However, commons theory admits that trust and reputation largely depend on face-to-face communication (Dietz, Ostrom andStern 2003, 1908). This requirement of face-to-face communication, in turn, relates closely to the need for a community to support the collective governance of certain resources, and the need for a practice of 'commoning' to support such communities. The central role of these elements (trust, reputation, face-to-face communication and a practice of 'commoning') suggests that commons-based institutions often function better when the spatial scale and the number of actors remains relatively limited (Araral 2014), as anonymity of the 5 Gibson, Ostrom and Ahn define 'levels' as ' [t]he units of analysis that are located at the same position on a scale.' They add that '[m]any conceptual scales contain levels that are ordered hierarchically, but not all levels are linked to one another in a hierarchical system.' (Gibson, Ostrom, and Ahn 2000, 218) 6 It has been suggested that commons scholarship and discourses have insufficiently taken the normative (including the political, social, economic and cultural) dimension of spatiality in the commons into account (Moss 2014). 7 Note, however, that Kaul et al. explicitly adopt a definition of 'global' which includes a sociological and temporal dimension, in that the goods in question should benefit or affect actors across social strata and across generations (Kaul, Grunberg, and Stern 1999b). from depletion and privatisation is, first and foremost, harmful to the local communities who dwell there. Hence, the primary authority over how to ensure the sustainable development in forest regions should lie with those groups, who have greater knowledge of the issues at stake (Ostrom 1994, 10-11). Even though both discourses emphasise the need to protect forested regions from privatisation, deforestation, degradation and illegal logging (thus confronting monocentric state authority or privatisation), their conception of the primary decision-making power differs greatly, not only in terms of the intended governance practices, but also on the understanding of who are the stake-holders involved.
The commons model has tended to consider that local populations and their plural mechanisms for governing fundamental resources (such as forests) should be the prime decision-making authorities over these issues (Ostrom et al. 1999, 281). In other words, governance should be bottom-up. Through an assessment of existing self-organised, longenduring governance practices at the local level, it argues in favour of allowing local communities to take the lead in the protection and promotion of fundamental goods, as the most appropriate way of dealing with various global issues (Ostrom et al. 1999, 279;Ostrom 1990, Ch. 3). The case of forest governance offers a very clear example of the opposing governance claims made by GPG and commons discourses (Dupuits and Pflieger 2017). Following an understanding of forests as commons, the local organisation of communities who live and depend on forest regions are not only the primary stake-holders and those who have more to lose from bad governance of this resource, but also the most effective and efficient authorities for ensuring its protection (Ostrom 1994, 20). Thus, governance should rely on a conic structure, which entrusts primary control and decision-making authority over resources to the communities who live and depend on them (Arnold 1998, Ch. 2). The local populations, as bearers of the primary authority over the governance of forests, are taken as active and direct participants in the protection and promotion of the goods in question.
On the other hand, the GPG discourse, as endorsed by Kaul and her co-authors, is, for the most part, a provision-oriented model. That is, its final objective is to ensure that a certain good -or protection from a certain 'bad' -is provided to the global population (Kaul, Grunberg and Stern 1999a, xxvi). The primary focus on the global population already determines the scale of stakeholders involved, thus pressing for a governance system that may best ensure provision to this wide public. The dual assumption of the selfishness of this plurality of actors and of the difficulty of achieving cooperative behaviour without strong and enforceable governance mechanisms (especially if global collective action problems are involved) demands, according to the GPG model, a top-down approach to governance (Quilligan 2012). Taking the case of forests as an example once again, the primary objective is to ensure the GPG of environmental protection (or avoiding the global public bad of greenhouse gases and global warming). Forests are, thus, intermediate goods of global interest due to their structural role in deterring global warming (Sandler 1997). Although, as an intermediate product, specific forests may be governed through different power regimes depending on particular circumstances; the overall governance of 'forests' (in a global sense) and their protection from deforestation, according to these discourses, should be a matter of primary authority of states and the global community as a whole (Kaul 2013, 5). The local communities who dwell in these regions, and the governance mechanisms they organise locally to protect their lands, are taken as intermediary agents whose power and decision-making authority over the spaces they inhabit depends on and is conditioned by the interests and outcome objectives that transnational entities have predetermined (Sand 2004). This top-down approach inherent in GPG governance models has been a source of concern due to the fact that its focus on provision can make it blind to issues of democratic legitimacy (Krisch 2014;Cogolati, Hamid and Vanstappen 2016, esp. 19-21).
GPG discourses have responded to these claims by appealing to the various levels at which and directions in which governance of GPG ought to take place, and to the potential implementation of more democratic involvement of plural global voices in their governance system (Kaul, Grunberg and Stern 1999c, 478-485;Schaffer 2012;Kaul 2013, 3). We submit, however, that even if these discourses defend the need to expand the democratic inclusion in the governance of GPG, this does not alter their top-down directive: defining what a GPG is, and how it should be governed in order to ensure effective provision, is determined by states and transnational authorities. It is mainly in the hand of a group of experts at the highest levels of governance who take the lead in defining both the objectives and avenues to achieve them, making local populations mostly passive recipients of goods provided in a top-down fashion (Boonen and Brando 2016, 141-147).
However, considering that there may be a democratic deficit in the governance model of GPGs does not imply that the commons discourse does not suffer from a similar ailment. The fact that commons theory implies a bottom-up directive in decision-making power should not lead us to conclude that the decision-making processes within commons is necessarily democratic. Communities have their own internal hierarchies, power-struggles and exclusion, and, despite the basic normative commitment of the commons discourse of empowering local communities and promoting self-governance, the particular processes and practices within each community determine its democratic credentials. 8

CONCLUDING REMARKS
Our comparison of the GPG and commons discourses has foregrounded differences rather than similarities. Both discourses seem to recognise similar challenges (e.g. a growing number of issues which are global in nature), but the solutions to these challenges are radically different.
The GPG discourse is concerned with effectively tackling issues of a global nature, through more inter-state cooperation and international institutions. Commons discourses, in contrast, reflect the ideal of self-governance of social movements and communities wary of market logic and state hierarchy. We summarise our main findings in the Table 1 below. We submit that the comparison between both discourses allows us to reflect more thoroughly upon the merits of the governance models they present. Indeed, the comparison has given us some insight into the strengths and weaknesses of each with the concept of power playing an important thread to link the tensions and conflicts between the two models. Two important and related points of contention arise: the discourses' democratic credentials and the role they assign to self-governance.
The GPG discourse emphasises the economic goals of effectiveness and efficiency. Indeed, it seems to be largely driven by experts embedded in international institutions. Although democratic amendments to its mainly output-driven rational have been proposed, these seem to come more as an afterthought, once it has already been decided which goods count as GPGs and who should provide them. These decisions are largely taken by a limited group of influential institutions, with participatory mechanisms providing social legitimacy. This democratic deficit becomes particularly apparent when compared to the active participation of communities in building up commons. The lack of awareness of GPGs discourses with democratic accountability in the past, and the strong concerns raised by critics of GPGs in this respect (Kirsch 2014), have led to recent work on the GPGs to tackle this issue more directly (i.e.

Brousseau et al. 2012). GPGs scholars have introduced many insights (directly or indirectly)
from the commons discourse in order to compensate for the discourse's hierarchical rationale.
For example, recent scholarly work has claimed the need to expand the scope of knowledge communities who should be involved at the stage of definition of what is a GPG, or the need to engage in a more profound process of collective learning between large-scale actors and local communities in order to ensure democratic legitimacy in the production and provision of GPGs (Brousseau and Dedeurwaerdere 2012).
Our comparison highlighted, as well, the central role that self-governance plays in commons discourses as opposed to the GPGs logic, which limits self-governance to a great extent. As explained above, the commons discourse adopts a conception of power which can be described as power-with. It is held that power originates from the autonomous cooperation within (and among) communities. This conception of power is also evident in the rejection of most forms of hierarchy (although certain manifestations of it may still remain). Commons discourse is presented as an alternative to state authorities and market pressures, but one can question whether it has the capacity (and strategy) to deal with the external pressures that emerge from larger and more powerful entities. Local commons are extremely vulnerable social and ecological systems, and their sustainability can be questioned due to their (seeming) incapacity to counter external threats. Recent additions to the commons-literature are starting to address