I. Introduction
Use of the concept of global public goods in the literature on international law has been increasing over the course of the past several years, following a trend in international development a decade earlier. The concept can indeed help to illuminate the role of and demands on international law in addressing a variety of global issues. But the concept is imprecise, and the term “public goods” is regularly used in the literature in two distinct ways. One way is to refer generally to issues that call for international legal and political cooperation. The other draws on economic logic, and refers to a much more specific set of circumstances. The broad understanding of public goods is a useful political call to international cooperation, but yields little in terms of guidance for the design and operation of international legal mechanisms with respect to particular issues. The narrower economic understanding, conversely, does yield specific prescriptions. The difficulty arises when the economic logic is applied indiscriminately to the broader understanding of public goods. Doing so can yield prescriptions that are misleading and counterproductive to effective cooperation.
Does this happen in the literature, and if so, does it matter? This article answers yes to both of these questions. While economic definitions of public goods speak of both non-rivalness and non-excludability, much of the literature on global public goods, including the literature in international law, operationalizes the definition by including anything that is either non-rival or non-excludable. This can result in the misdiagnosis of the legal structures needed to effectively enhance international cooperation and the creation of legal institutions that are inappropriate, in issues ranging from human rights to fisheries management to geoengineering. We argue that while creating a broad category of global public goods may be politically useful in generating international cooperation, it can be analytically problematic in designing international law. For example, many issues in international cooperation, including but not limited to most environmental issues, can be more usefully characterized analytically as common pool resources (CPRs) rather than public goods. Other issues have both public and CPR (as well as private) aspects, and the design of appropriate international law needs to take all of these aspects into account. Using the concept of public goods in its strict economic definition as a basis for nuanced examination of the specific structures of issues demanding international cooperation, and allowing for the distinct category of CPRs, yields a more precise understanding of the legal structures necessary for efficient and effective international cooperation on those issues.
This article begins with a brief review of the literature on global public goods, a discussion of the definitional issues touched on in this introduction, and an argument that many issues in international law are better understood as CPRs than public goods. It then makes three sets of arguments that situate analysis drawing on the economic definition of the concept of public goods in the study of international law. The first focuses on a key difference between public goods and CPR issues—that the former is about the production of a good, and the latter about consumption. This affects how international legal mechanisms such as enforcement should be designed; public goods mechanisms should focus on encouraging those most willing to produce, whereas CPR mechanisms must focus on discouraging those most able to consume. The second set looks at ways in which global public goods and CPRs interact differently with questions of exclusion and legitimacy, since creating mechanisms of exclusion for a public good creates a club good, whereas doing so for a CPR creates a private good. The third steps back from the details of rivalness and excludability and asks what constitutes a good in the first place, and argues that effective legal mechanisms for determining what kind of good something is must be more inclusive than those for generating or managing that good. In the final section, we offer geoengineering as an example of the ramifications of these issues.
II. Public Goods and International Law
The use of public goods as an analytic concept is gaining increased currency in the discipline; the majority of this literature discusses the tools with which international law can facilitate the provision of global public goods. For example, in 2011 the American and European Societies of International Law created
a research forum to explore whether and how the co-existence, interaction, and antagonisms of a plurality of legal orders (international law, domestic law, European Union law, regimes established by private actors) and their driving agents (regulators, contract-makers, and courts and tribunals) contribute to creating and maintaining global public goods.Footnote 1
This follows on the widespread adoption of the concept in the field of development studies.Footnote 2
The international law literature discusses the inadequacies of the current international tools and principles for the provision of global public goods, understood as those public goods that are associated with global issues. Scholars in the field focus on various aspects of the underproduction of global public goods, such as the role that private international actors are starting to play in supplying them,Footnote 3 how to finance them,Footnote 4 and validating the legitimacy of those who provide them.Footnote 5 The literature also highlights ways in which current practices in international law are inadequate for the effective provision of global public goods. Nico Krisch, for example, offers problems regarding supply of these goods as one of the reasons why the use of consent- or consensus-based decision-making in international law is anachronistic.Footnote 6 Other scholars, meanwhile, argue for increased attention to issues of consensus and consent in discussions of global public goods provision. Ernst-Ulrich Petersmann, for example, argues that having legitimate constitutional and judicial processes in place that are broadly supported by the affected citizenry is necessary for the democratic provision of global public goods.Footnote 7 Neil Walker argues that both the ideas of human rights and public goods face the same type of challenge, which is reconciling political authority with political morality.Footnote 8
The literature on global public goods in international law suggests that scholars increasingly use the term to cover a broad range of international law that deals with issues of environment, trade, financial stability, health care, poverty reduction, and culture preservation across states’ borders. This literature, however, is better at identifying obstacles to, than mechanisms for, effective international law on these issues. We argue that this is because the concept of public goods is being used in an analytically imprecise fashion in much of this literature. Casting an issue as one of public goods focuses discussion around provision of a good, but many of the key challenges facing international law in many of these issues has nothing to do with provision per se. A disaggregation of the concept of public goods allows us to see the relationships among provision and consumption in these issues.
III. Public Goods and Common Pool Resources
Many treatments of public goods in the international law literature that explicitly define the concept refer to its economic components without necessary applying them fully. In introducing the special issue of the European Journal of International Law on the topic, for example, Fabrizio Cafaggi and David Caron define a public good “as one that is characterized by non-rivalry (anyone can use a good without diminishing its availability to others) and non-excludability (no one can be excluded from using the good).”Footnote 9 This is an accurate statement of the economic concept.Footnote 10 But few treatments move beyond this basic shared definition to actually look at its component parts, how they are different from each other as social scientific concepts, and what happens when they vary differentially.
Rivalness and excludability are different in that they describe different characteristics of a good; however, they are also different in that the characteristics are of different kinds.Footnote 11 Rivalness is, to a large extent, a feature of the physical characteristics of a good, and of how individuals value those characteristics. Firewood is rival because the process of burning it physically uses it up; a radio broadcast is non-rival because tuning a radio to a broadcast has no physical effect on the ability of other radios to tune in. Rivalness is not just a matter of physical characteristics, however. A whale is rival as meat, but non-rival to the extent that we value it as a fellow creature. To the extent that some people, or, as expressed by national law, some countries, value it as one and some value it as the other, the whale is both rival and non-rival at the same time.
Excludability is a different kind of characteristic in that it is, to a significant extent, a legal construct rather than inherent in the good or the way in which we value it. While it is the case that some goods are easier to exclude in practice than others, goods are in large part excludable by law rather than intrinsically. The difference between a public park and a private park is ownership in law. The quality of the fence surrounding the park affects the practical ease of exclusion, but not the basic characteristic. Some goods may in practice be prohibitively difficult or costly to exclude, but relatively few—a radio broadcast is actually quite easy to exclude, for example through the use of subscription codes, while a lighthouse beacon is more difficult. To use an example germane to international law, ocean resources between three and two hundred nautical miles from land were non-excludable until states decided to claim, and to recognize, exclusive economic zones.Footnote 12 Those same resources then became excludable, and remain so in principle even if many states do not have the physical capability to enforce that exclusion in practice.
Since rivalness and excludability do not necessarily co-vary, a dichotomous reading of the two characteristics yields four ideal types rather than two categories of goods (see table below). A rival and excludable good is a private good, the provision of which should not require international cooperation (although legal clarity about what is and is not excludable, which is a prerequisite of a well-functioning market at either the domestic or international level, is itself a public good). A non-rival and non-excludable good, as already noted, is a public good. But there are two other possibilities as well. A good that is excludable but non-rival is generally referred to as a club good. Since free riders can be excluded from club goods, issues in the provision of these goods are more likely to be about the details of the rules of the club and the distribution of its benefits, rather than who will participate in provision.Footnote 13 An example of this kind of club good at the international level is international trade rules.Footnote 14
The fourth kind of good yielded by this categorization is rival but non-excludable—these are CPRs. The fact that CPRs are non-excludable means that they share with public goods a free rider problem. The fact that they are rival, however, means that the politics and law of providing them are quite different, as will be shown below. It is our contention that many aspects of international cooperation that are often looked at as public goods, particularly within the issue area of international environmental politics, are better understood as CPRs.
To the extent that treatments of the concept of public goods in the international law literature do complicate the concept beyond a public/private good dichotomy, they often do so in one of two ways that fail to address the useful analytic distinction between public goods and CPRs. The first is to speak of imperfect or impure public goods, putting individual goods somewhere on a one-dimensional scale between private and public. The phrase “impure public good” is often used in the international law literature to refer to goods that are partially excludable (and/or partially rival).Footnote 15 But this usage obscures the difference between what one might call a “pure” impure public good, one that is non-rival and non-excludable in similar measure, and goods that fit better into the category of CPR. For example, an urban public park is often both partially rival and partially excludable. It is partially rival because after a certain level of usage there is a crowding-out effect, in which more users can undermine the park experience of existing users. It is partially excludable because it can be fenced off and closed to users, for example at night. An international marine protected area is also partially excludable, in a different way—exclusion is legal, through a multilateral agreement, rather than physical, by means of a fence. But it is far more rival than the park, in that it only takes one large bottom-trawler to completely undermine its utility as a protected area. The challenges of managing an international marine protected area are, therefore, more like those of a common pool resource than those of a public good.Footnote 16
The second is to develop categorizations that do not draw clearly on the basic definition of public goods.Footnote 17 The above-mentioned special issue of the European Journal of International Law, for example, draws heavily on a tripartite categorization of public goods developed in the work of Scott Barrett.Footnote 18 In his categorization, single best effort public goods are those that can be provided by a single actor, weakest link public goods are those that can be undermined by a single weak or uncooperative actor, and aggregate effort public goods require a broad cooperative effort for effective provision. The tripartite categorization is an example of obscuring the difference between impure public goods and common pool resources. What Barrett calls “single best effort” goods (his examples include peacekeeping and knowledge) are public goods in the narrow definition provided here; they are both non-rival and non-excludable to some degree. Because they are not excludable, they generate classic collective action problems. Because they are not rival, the cost to providers of the good of supplying it to everyone is no higher than the cost of supplying it to themselves. Therefore, if a single country derives more benefit from a global public good than the cost of providing it unilaterally, then it will likely provide it, even absent cooperation.
Many international public goods will not fit into this category, in which unilateral action makes sense for a single country (usually the biggest player in the relevant context). It may well be that provision only makes economic sense to a group of countries; a group for which unrequited public goods provision makes sense that is a subset of a larger whole is called a k-group in the relevant economics and political science literature.Footnote 19 With international public goods, this usually means the biggest and richest countries. Examples include international security, in which the permanent members of the United Nations Security Council can reasonably be seen as a k-group, and the research done by the Organization for Economic Cooperation and Development, which includes only developed countries as members. Other things being equal, more concentrated provision of public goods is likely to be more effective—the larger the k-group, the greater both the transaction costs and the free rider problem within it. These goods can be provided effectively without the participation of states less interested in the good. With public goods, free riders do not contribute to provision, but at the same time do not act as spoilers.
Barrett's other two categories are “weakest link” and “aggregate effort.” Weakest link goods are those for which cooperation can be undermined by a single spoiler, such as disease eradication. Aggregate effort goods are those for which the participation of all states is required for effective provision—his examples include climate change mitigation and ozone layer protection (both, interestingly, fit into the category of air pollution prevention).Footnote 20 These two categories are not as distinct as their nomenclature indicates—it is in the nature of aggregate efforts that they can be undermined by weak links. And even in the case of binary goods such as disease eradication, efforts short of total success will still yield benefits, meaning that a spoiler cannot completely undermine the good, and aggregate effort matters. Furthermore, from the perspective of international cooperation, most of the examples used for these two categories can usefully be thought of as CPRs rather than public goods,Footnote 21 and, as we will discuss below, thinking of them as CPRs yields a more nuanced analysis than as types of public goods.
IV. CPRs and International Cooperation
To say that the issues can usefully be thought of as CPRs may sound at first a little imprecise or slippery—one might assume that they either are or are not. But the situation is in fact more complicated than that, because there is, in many cases, a difference between what we think of as the proximate good that we are concerned with, and what is the proximate focus of international cooperation. Take as an example ozone layer protection. The ozone layer itself is close to being a pure public good; we are all shielded by it equally, there is no partial rivalness of a crowding-out effect, and it is difficult to exclude some people from the benefits that others get at the same time.
Having said this, the proximate focus of international cooperation is not the provision of an ozone layer per se. The ozone layer is provided by nature. The proximate focus of international cooperation is limiting the production and use of ozone-depleting chemicals such as chlorofluorocarbons (CFCs).Footnote 22 In other words, cooperation is about the use of the ozone layer as a pollution sink for a category of chemicals, rather than using it as protection against ultraviolet radiation from the sun. And pollution sinks (except for those that keep the pollution entirely within privately owned territory, without any threat of leakage) are not public goods; they are CPRs.Footnote 23 Like public goods, anyone can use them, in this case by disposing of ozone-depleting substances. But unlike public goods, pollution sinks are rival. The more CFCs one actor dumps into the atmosphere, the less others can dump without giving us all melanoma. International law to protect the ozone layer therefore must focus on the CPR of the pollution sink, rather than the public good of the ozone layer itself.
How widespread are CPRs among the issues calling for mechanisms of international law? More so than might be obvious at first glance. International resource use issues are the most obvious ones, and can be seen straightforwardly as CPRs. Resources like fisheries and seabed mineral deposits are clearly rival, and if they are to be found outside of exclusive economic zones (EEZs), they are, absent treaties to the contrary, non-excludable. It is also the case, though, that pollution issues in the international commons are CPRs, to the extent that the focus of cooperative efforts is the use of the global commons (including the atmosphere and the oceans) as pollution sinks. This includes both ozone depletion and climate change—the politics are about how much pollution can be emitted, not about access to the climate or the ozone layer in general. It also includes issues of ocean pollution. In all of these cases, rules and procedures are about who gets to emit how much pollution, or, in other words, who gets to use up how much of the rival good of the pollution sink, and who will pay the costs of mitigating that use.Footnote 24
Most international environmental issues, in other words, are CPR issues (the major exception to this generalization is issues that are strongly directional, involving rivers or air pollution such as sulfur dioxide that is carried downwind—these are best seen as disputes regarding property rights over the externalities of private goodsFootnote 25 ). One set of international environmental issues that seems at first glance to not fit into this resource use/pollution dichotomy is species and biodiversity protection. While this issue as a general category may be difficult to categorize, specific problems of species and biodiversity protection are in fact generally about either the use of species as a resource or the protection of habitats from various kinds of pollution (including expropriation for human use, which from an environmental protection perspective can be seen as a kind of pollution). For example, the Convention on International Trade in Endangered Species is focused on specific species as targets of resource exploitation, while the Ramsar Convention is designed to improve cooperation in the prevention of pollution and encroachment of a particular kind of habitat.Footnote 26
What of goods that fit into Barrett's “weakest link” category? In order to fit fully into this category a good needs to be fully binary, in the sense that it is either fully provided or not provided at all, with no possibility of partial provision (otherwise a single spoiler would not be able to fully undermine cooperation). Disease eradication arguably fits into this category of binary good, although as noted above there are benefits to disease control short of eradication. Vessel reflagging does not, and one can in fact make the argument that vessel flagging should not be thought of as a good at all, but rather as a regulatory mechanism designed to manage other goods, like fish or pollution control.Footnote 27 And while vessel reflagging is indeed an obstacle to achieving these goods through international cooperation, casting it as a binary good is misleading. International efforts to manage international fisheries and marine pollution from ships have developed several mechanisms for addressing the problem of reflagging. These are not perfect, but are good enough that a single weak link that reflags is unlikely to negate international cooperation on these issues altogether.Footnote 28 And these issues themselves, for the reasons discussed above, are best seen as CPRs.
A final question in this specific context is whether there are international goods other than environmental goods that are best characterized as CPRs. The answer is that this characterization is appropriate for issues that we think of as relating to a global commons. This includes environmental commons, such as the atmosphere and oceans. It also includes other goods provided by nature that we do not necessarily think of in the context of environmental protection, such as electromagnetic bandwidth.Footnote 29 Another category of goods includes cultural commons, sometimes referred to as “global cultural goods.”Footnote 30 As is the case with the ozone layer example, we can reasonably think of the maintenance of cultural goods as public goods provision. But direct threats to cultural goods, much like direct threats to species habitat, introduce a CPR element, because damage done to cultural goods is rival.
V. Production, Cooperation, and Enforcement
Two arguments that are central to the claims being developed in this article build on these observations about CPRs and international cooperation and law. The first is that in discussion of international public goods and CPRs there is an important distinction to be made between the production and consumption of goods. The second is that we can usefully look at different aspects of international cooperation about a given issue as involving different kinds of goods.
Production and Consumption
A key difference between issues in international law that have public goods characteristics and those that have CPR characteristics has already been alluded to above: public goods are produced by people, either by single providers or by cooperative effort. CPRs are provided by nature, and consumed by people. As such, the legal and political structures created to deal with public goods must be designed to manage the process of producing the goods in order to be effective, whereas the structures created to deal with CPRs must be designed to manage the process of consuming the resources. This conceptual distinction is useful in thinking about the different legal and political challenges in cooperation to manage these goods effectively, in part because it shows why spoilers are a problem for CPRs but not for public goods.
Governance of both kinds of goods is plagued by the problem of free riders. But the problem of free riders is different for each. For public goods, free riders can increase the cost of provision to cooperators, and may result in the underprovision of the good. But free riding cannot undermine the value of the good once provided.Footnote 31 For example, Canadian free-riding on U.S. provision of collective defense can increase the cost burden to the United States of providing that defense, but Canadian free riding does not directly undermine the ability of the United States to provide the level of collective defense that it chooses.Footnote 32 Cooperation can proceed effectively without bringing in free riders, as long as providers are willing to bear extra costs. The damage a free rider can do is limited by a zero boundary—if it contributes nothing, it can do no more harm.
With CPRs, however, free riders can be more pernicious, because they are consuming too much rather than producing too little. The key difference is that overconsumption does not have the equivalent of the zero boundary of underproduction. With CPRs, free riders can (in some cases) increase consumption as much as cooperators decrease their consumption, leaving the CPR in the same shape as before cooperation. All that has changed is a shift in consumption from cooperators to free riders. This means that with CPRs, free riders can act as spoilers of cooperation to the extent that they can consume the CPR.
Note that this is different from the role of spoilers in the case of dichotomous goods in “weakest link” cooperation. It need not be the case that a single free rider negates cooperation—spoilers can partially, as well as fully, undermine cooperation. Also, the language of “weakest link” can be misleading. The free riders that do the most damage to international cooperation over CPRs are less likely to be states that do not have the capability or resources to fulfill their international obligations. Rather, they are likely to be those states that have the greatest ability to use up the resource.Footnote 33 For example, in the case of climate change mitigation, since China is by far the largest emitter of carbon dioxide at the moment, China can act in effect as a spoiler for any meaningful attempt to limit emissions in which it does not participate. Botswana, however, cannot act as a spoiler because it does not have the capacity to emit enough carbon dioxide to make a difference on a global scale. More broadly, the greater the ability to use up resources or to pollute at large scale, the greater the extent to which a country can act as a spoiler to international cooperation over CPRs.Footnote 34 The fact that free riders can act as spoilers, and can actively undermine cooperation to manage CPRs, makes power relationships in negotiations on CPR issues different from negotiations on public goods issues. Since spoilers can undermine cooperation, they need to be brought on board in a way that is not true for free riders on public goods. In other words, with public goods free riders can be ignored if cooperators are willing to bear the costs of provision. With CPRs they must be induced (or coerced) to participate for legal mechanisms to be effective.
This distinction makes the dynamics of CPR cooperation fundamentally different from the dynamics of public goods cooperation, because it affects the likely distribution of costs and benefits associated with the operative legal mechanism. Any such mechanism is in part about identifying the extent to which a good will be provided or exploited, but it is, in part, also about distributions of the resultant costs and benefits. These costs and benefits are sometimes matters of scale, such as who can emit how much pollution or catch how many fish, or how much collective security will be provided. But they are also sometimes matters of kind. For example, the provision of international security can take different forms, and have different political emphases; fishing can use different equipment, with different environmental effects. With public goods, those providers willing to bear the greatest costs are likely to have the greatest say in defining the form of the good and the content of law regulating it. With CPRs, those potential free riders most able to act as effective spoilers are likely to have the greatest say in these matters, as well as in questions of who gets how much of the resource.
Cooperation and Enforcement
The observation that power dynamics in the creation of international law regarding management of CPRs are different from those of public goods is important to the extent that the goods in need of such management have the characteristics of CPRs. But, as we saw above, goods do not necessarily have only one set of characteristics. The ozone layer as a source of protection from UV radiation is reasonably understood as a public good, but the ozone layer as a pollution sink for a particular category of chemicals is better understood as a CPR. From the perspective of negotiating cooperative decreases in the emission of ozone-depleting substances, the latter framing is the more relevant.
But the negotiation of decreases is not the only relevant aspect of effective international law. Monitoring and enforcement of agreements once they have come into effect can also be key to managing CPRs. And monitoring and enforcement can themselves have public goods, rather than CPR, characteristics.Footnote 35 The provision of individual monitors, the maintenance of GPS satellites, and the gathering of intelligence on weapons proliferation are all examples of monitoring efforts that are non-rival and can be, in the context of a cooperative agreement, non-excluded. Similar observations can be made about efforts to effect and coordinate enforcement, such as Australian efforts to coordinate internationally the port exclusion of vessels fishing in the southern oceans in contravention of Convention for the Conservation of Antarctic Marine Living Resources rules.Footnote 36
This distinction between the CPR characteristics of cooperation to protect the global commons and the public goods characteristics of elements of the monitoring and enforcement of that cooperation helps to explain why the latter is more prone to unilateralism than the former. Since the logic of public goods as applied to international cooperation demands a focus on provision rather than on consumption, the distinction also tells us that in the monitoring and enforcement of international legal mechanisms to protect the global commons, we should focus on who is willing to provide public goods rather than on the free riders. This entails finding out if unilateral provision is a possibility, and if not, whether a viable k-group can be established. In creating regulation of the use of the commons, conversely, a concern with addressing free riders is key to cooperation. Creating effective international commons law, in other words, calls for different patterns of participation than enforcing it.
Understanding the distinction as one between CPRs and public goods aspects, rather than as different kinds of public goods, is more intuitively useful than understanding it as a distinction between aggregate effort public goods and single best effort public goods for two reasons. The first is that it makes it easier to discuss different aspects of cooperation about a given issue. When discussing, for example, climate change law, asking what parts of the cooperative process are rival and what are not clearly identifies which elements are public and which are common pool. The second is that the language of aggregate effort and single best effort is not as descriptively accurate as a language that focuses on the distinction between provision and consumption.
VI. Excludability, Legitimacy, and Law
The argument to this point has focused on identifying the differences between the logics of public goods and CPRs and delineating in what way different issues fit into each category. This section of the article looks more specifically at ways in which this difference matters for questions of international law. It considers two specific issues concerning the differential effects of public goods and CPR cooperation: the role and effects of exclusion, and the question of legitimacy.
Exclusion
Excludability is, to a large degree, a creation of law rather than inherent in a good, and can therefore be changed in law. Making a good legally excludable where it had previously been non-excludable changes a key characteristic of the good. The result is significantly different for public goods and CPRs. Public goods, when made excludable, become club goods (excludable and non-rival). CPRs, when made excludable, become private goods (rival and excludable). One practical effect of this difference is that international public goods when made excludable remain cooperative goods, because there are strong incentives to cooperate in the provision of club goods. But CPRs, in becoming private goods, are removed from the realm of international cooperation and become matters of purely domestic concern.
The creation of mechanisms of excludability, otherwise known as privatization, is one of the key strategies offered in the literature on CPRs for addressing overconsumption.Footnote 37 The logic of privatization is that it allows the owner of the newly privatized good to internalize all of the benefits associated with good stewardship of a resource, increasing the incentive to manage the resource well. There are arguments in this literature against privatization, including normative arguments against taking goods from common possession and, by entrusting them excludably to some actors, necessarily at the same time denying them to others.Footnote 38 There is also the broader Rousseauvian question of whether a cumulation of private goods equals the public good.
This latter concern can be seen in perhaps the biggest example of privatization of an international commons in the modern era, the creation of EEZs and their formalization in international law as part of the UN Convention on the Law of the Sea in 1982. EEZs were supposed to have the effect of making resource management, and in particular fisheries management, more sustainable on continental shelves outside of territorial waters.Footnote 39 In some cases this worked, by giving individual governments both the authority to regulate the resource and a vested interest in maintaining its long-term health.Footnote 40 But in many cases, the creation of EEZs led to increases in levels of exploitation of resources (particularly fisheries resources), the opposite of what was expected. Many countries responded to their new level of control over fisheries by expanding their fleets so as to be able to maximize domestic industry revenue, while excluding vessels of other nationalities.Footnote 41 The net effect in the medium term of this experiment in large-scale privatization was to increase the capacity of the global industry to overfish, rather than to generate more sustainable management.Footnote 42
When public goods are made excludable, they are not privatized per se; rather, they become club goods. As an example, unilateral free trade can be seen as a public good. Discussions in the United Kingdom in the mid-nineteenth century about eliminating the corn laws and freeing up trade more generally were clear on this point—free traders’ first preference was for everyone to free up trade, but in the absence of reciprocation they still argued that all countries, including the UK, would be better off if the UK unilaterally reduced or eliminated tariffs. Making tariff reduction excludable by creating a club within which the rules of nondiscrimination apply, leaving members free to keep tariffs high on non-members, changes the public good into a club good.Footnote 43 This is essentially what the World Trade Organization does. This sort of arrangement differs from both public good and CPR cooperation in that the carrot of membership can be used to gain agreement to the rules from potential members, and the stick of exclusion as punishment can be used to punish breaches of the rules, but also differs from a privatization intended to dispense with the need for continuing cooperation.
Clubs can also be created to support legal mechanisms for cooperation with respect to related public goods and CPRs. The point of exclusion for such clubs in international law is often market access. So, for example, the Financial Action Task Force (FATF) can recommend that member states restrict access by non-compliers to their financial markets;Footnote 44 the Kimberley Process keeps diamonds mined outside of its rules from the markets of signatory states;Footnote 45 and ships that do not comply with international safety, labor, and environmental standards can be kept out of major ports through both public and private enforcement mechanisms.Footnote 46 Both the diamond and shipping cases are examples of clubs being created and enforced by a hybrid of state and nonstate actors.
Legitimacy
This discussion of the differential role of excludability in different types of goods highlights one way in which focusing on the characteristics of goods, rather than using of public goods as a broad and generic category, can be analytically useful. Another way in which doing so can be helpful in thinking about the role of international law and international cooperation in governance has to do with legitimacy. Issues of legitimacy are distinct for different kinds of goods, and vary according to the characteristics of goods, in a way that is generally not captured by the literature on global public goods and international law. For example, Daniel Bodansky provides a useful overview of questions of legitimacy and law in the provision of global public goods.Footnote 47 In doing so, he addresses the different specific questions generated by Barrett's three different technologies for the production of global public goods. Best effort goods, he argues, are less likely to generate legitimacy questions, whereas the coercion that may be necessary to provide aggregate effort goods is more likely to generate legitimacy questions. With weakest-link goods, meanwhile, it depends. If the weak link needs to be coerced to participate, then there will likely be legitimacy issues, but if it needs help to participate, then likely not. We argue that these legitimacy issues are analytically clearer if approached through the lens of the characteristics of goods rather than the lens of different production technologies.
The starting point for such an analysis is to look at the legitimacy effects of the component characteristics of the different kinds of goods, rivalness, and excludability. Legitimacy concerns arise in two ways in this context. The first is when changes to a status quo have negative externalities on third parties (in this case states) that are not addressed by the parties that generate the externalities.Footnote 48 The second is when actors (again, in this case states) are required to do things, through the creation of new law or through other means of coercion, that they would not otherwise do and have not consented to doing.Footnote 49 Agreements that allow third parties to continue behaving as they have been, or a status quo that does not result in negative externalities that affect them, should not generate legitimacy concerns.
As goods move along the continuum from non-rival to rival, we should expect legitimacy concerns to increase, other things being equal. Consumption of a non-rival good is unlikely to create legitimacy concerns because it has no effect on anyone else's ability to consume that good. At the other end of the spectrum, consuming a fully rival good creates the negative externality of denying that good to all other parties. This creates legitimacy issues to the extent that some or all of those other parties had a reasonable expectation of access to the good. The key concern here is about fairness in distributing losses across consumers.Footnote 50
As goods move along the continuum from non-excludable to excludable, the ability of the producer of the good to retain consumption benefits from the production of the good increases. Being located on the excludable end of the spectrum does not in itself generally create questions of legitimacy—a widget maker is generally considered within its rights to retain the proceeds from the sale of its widgets. The questions arise when goods move along this spectrum, or when different parties disagree about where on the spectrum a good should be.Footnote 51 The legitimacy concern in this case is about what extra benefits are due to producers. An example of such a concern can be found with intellectual property rights, which are a form of legal exclusion from benefits. A change in such rights (such as an increase or decrease in the length of patent protection) inevitably has distributional effects, and as such may well be seen by the distributional loser, in the absence of an accepted process for changing the law, as illegitimate.
Pure public goods, by this analysis, should not generate legitimacy concerns. Consumers of the good do not affect others’ ability to consume, and producers do not capture any special benefits from production. Examples here might include basic science, or deflecting an incoming asteroid. There are of course no pure public goods, inasmuch as goods have different aspects. In both of these examples the producer of the public good generates private gains as well, in maintaining an effective national science infrastructure or in developing technologies that have uses beyond asteroid deflection.Footnote 52 But both examples are public enough that the legitimacy concerns are likely to be minimal.
Club goods should not generate consumption-related legitimacy issues, because consumption is non-rival. But they can generate production-related legitimacy issues, to the extent that the creation of the club in question favors some interests and excludes others. This is particularly true of clubs that show positive returns to scale, meaning clubs that provide greater gains the more members they have. While the rules themselves may favor everyone, whoever is able to create the rules for the predominant club (whether they did so because they were the first to create the club, or because size or willingness to produce the good allows them greater weight over the rules of the club) benefits most.Footnote 53 This can raise legitimacy concerns about the disproportionate role of some states in the creation of international rules that affect all states.
Examples here include cooperation over international security, trade rules, or product standards. In the case of international security, those states able and willing to spend most on military capabilities will have the greatest say over what constitutes a threat to international security.Footnote 54 In the case of trade rules and international product standards, most states are generally better off with common rules and standards than with none, but the details of those rules and standards will necessarily favor some actors over others. The existence of these rules and standards, as well as the creation of new ones, therefore, can have the effect of entrenching existing inequalities.Footnote 55
International cooperation over the management of CPRs can generate legitimacy issues both in their consumption and their production. With respect to consumption, effective cooperation requires that all major users of the resource participate in a management scheme; otherwise, non-participants can simply use up more of the resource in response to the restraint shown by cooperators. Thus, to be effective, rules need to be applied to all major users, whether or not they consent to them. To use the example of carbon emissions, rules that do not apply to the single largest emitter, China, will not be particularly effective at solving the problem. But China can legitimately claim “common but differentiated responsibilities” as a developing country.Footnote 56 The tension between legitimate claims regarding the need to reduce emissions and legitimate claims for differentiated responsibilities hobbled climate change cooperation for decades.Footnote 57 To use as another example international fisheries cooperation, recent efforts by some Regional Fisheries Management Organizations to exclude fish not caught within their rules from major international markets has the effect of forcing their management rules on states that have not consented to them.Footnote 58
At the same time, many attempts to deal with international CPRs involve some form of privatization, intended to make them into private goods in an international sense (i.e. in the sense that they are private to one country; however, they are then managed within that country). But privatization has the effect of taking something that was communally owned, and depriving some states of that ownership. Examples include the creation of EEZs, the distribution of rights to the electromagnetic spectrum, and rules limiting the right to pollute the atmosphere or oceans. This raises questions about the legitimacy of depriving states of rights to a commons without their consent.Footnote 59
VII. Goods and Commons
A prior step to legitimate law about a public good or CPR is determining what constitutes a good or a commons in the first place. The literature generally assumes we know these things when we see them, but that is often not the case. In economic parlance, a good is something with positive utility, something that, other things being equal, we want. This is, in general, a definitional category that does not trouble economists much; if an individual is willing to engage a market to get something, it is a good for that person, and if not, not. But in the context of a discussion of global public goods and commons this definition raises the question of what happens when the public disagrees about whether something is a good. We cannot make effective law about global public goods unless we have a legitimate process for figuring out what those goods are.
There are a variety of things or actions that some actors consider a good and others do not. At the international level, for example, we may disagree about whether certain kinds of international banking rules constitute a public good (because they limit the likelihood of financial crisis) or a public bad (because they make the international allocation of capital less efficient).Footnote 60 In terms of issue structure, public bads are not a mirror image of public goods. Rather, they often have the characteristics of common pool resource problems. A poorly executed regulation reduces the degree to which the public good of international financial stability is robust against other activities that can undermine it, and thereby in a way uses up some of the common resource of international financial stability.
The observation that there is often some disagreement on the question of what constitutes a good raises problems for determining whether a particular thing is to be collectively considered a global good (whether public or commons) when there is no global consensus. A philosopher might answer the question of how to make such a determination by deducing an idea of the good from philosophical principles, for example of a general will or a veil of ignorance.Footnote 61 An economist might try to measure whether people on average value a good, while a political scientist might ask who has the power to impose a particular vision of the good. A legal approach to answering the question should address questions of process—something would be considered a global public good, or part of the global commons, if determined to be so through some kind of legal process.Footnote 62 Such a process can generate legitimacy in the determination of what constitutes a good, or, in other words, which public goals international mechanisms ought to address, even when consensus is absent.
Many existing processes in international law are designed to do this. However, these processes have both different levels of authority and quite different standards of agreement. For example, in the United Nations Security Council, a mere nine states can determine whether something constitutes a threat to international security and therefore warrants authorizing the public good of an Article VII response. The United Nations General Assembly can identify the global good by simple majority vote, but cannot generate hard law to provide or ensure it. The World Trade Organization assumes that consensus among member states is required for determining whether an international trade rule is in the public good.
A complicating factor in creating international legal institutions in this context is that different legal mechanisms are effective for generating processes for deciding whether or not something is a global public good, creating public goods once that decision has been made, and protecting the commons from global public bads. An institution designed to create a public good would focus on creating a k-group and holding it together, most likely by creating some kind of associated club. An institution designed to protect the commons from a public bad would need to focus on enforcement against those most likely to cheat and continue to use up the rival resource. Neither of these institutions would need to be broadly inclusive. But an institution designed to legitimate a process for determining what constitutes a good in the first place would need to place a premium on inclusivity and deliberation, even if at the expense of an ability to privatize or make enforceable hard law.
VIII. Geoengineering
A useful example of this institutional design dilemma can be found in the issue of how to govern efforts to address climate change through geoengineering,Footnote 63 and particularly through what is called solar radiation management, or SRM. This refers to any attempt to lower the global temperature by reducing the amount of sunlight that makes it to the surface of the planet. The technology of SRM ranges from the available and cheap, such as injecting aerosols or sulfur into the atmosphere, to the not-yet-invented and potentially expensive, such as large mirrors in space. SRM has both proponents and detractors; the details of that debate are not germane to the argument here. For proponents, the deployment of SRM would be a public good to combat climate change.Footnote 64 For detractors, such a deployment is too uncertain, and might do more harm than good.Footnote 65 For them, preventing such a deployment helps to protect the common pool resource of an unengineered atmosphere. The creation of new knowledge, both about the side effects of SRM and of the severity of the effects of climate change, could potentially move actors from one group to the other.
If SRM is understood as a good, then the role of a governance mechanism would be to coordinate efforts among those states interested in and capable of constructing, or contributing to the construction of, SRM systems. This mechanism would work best if it included only the relevant k-group and focused on what Gregory Shaffer calls global administrative law.Footnote 66 Free riders would not be a problem, because the k-group could provide the good without them. Free riders might introduce minor inefficiencies, but would not constitute a threat to cooperation, and could therefore reasonably be ignored. If SRM is understood as a bad, however, the role of a governance mechanism would be specifically to prevent free riders (which would in this case be defined as actors who undertook SRM). It would therefore need to be focused on enforcing a prohibition on deployment of SRM, whether this deployment happened within or outside of the membership of the governance mechanism. This is a key difference between public goods and CPR cooperation—in the former free riders are a nuisance, in the latter they are a critical threat. Finally, if it is unclear whether SRM is a good or a bad, a governance mechanism would need to focus on generating a mechanism for deciding. This focus suggests an organization that is as inclusive as possible, and that cannot reasonably be seen as an effort by powerful countries to impose a decision on the rest of the world.
Ian Lloyd and Michael Oppenheimer argue in favor of a governance mechanism that does all three. They propose starting with a k-group, to start the process of getting international legal structures for SRM in place, and gradually expanding and democratizing it as the science both of SRM and of climate change becomes clearer.Footnote 67 Such a mechanism would at various stages match all of the needs listed in the previous paragraph for dealing with SRM as a good, a bad, and an unknown. But it gets the sequence wrong, and could therefore end up not being effective at dealing with any of them. Lloyd and Oppenheimer argue that their mechanism would serve at first to prevent deployment of SRM, and then to develop a legitimate process to decide whether to allow and enable it. But the original k-group design is inappropriate to the preventive phase, because a country that wants to deploy SRM can simply not join the group, and would thereby be under no obligation not to deploy. This in turn obviates the decision phase. The sort of legal pluralist structure most suited to the task of legitimate decision-making about whether SRM constitutes a good or a bad would be necessary as a first step, but is unnecessary for the public goods provision stage, when they propose creating it.
This example illustrates the importance of problem structure definition, and of matching problem structure to legal structure. Beginning with an assumption that one is dealing with some kind of generic public good can lead to sloppy thinking about what kind of structure is needed or useful. The sort of legal mechanisms likely to make public goods provision more efficient is quite different from the sort likely to protect a commons effectively. Similarly, a legal mechanism intended to provide a public good or protect a commons should have a fundamentally different design from a mechanism developed to legitimately determine whether something is a good or a commons in the first place.
IX. Conclusion
Using the language of public goods precisely can illuminate the role and effects of international law in issues of international cooperation more clearly than the broader definition of public goods often seen in international law journals. Using the language precisely highlights the difference between the need to produce public goods and to prevent the consumption of commons goods, and the concomitant differences in who needs to be included in the relevant international legal mechanisms. It allows us to distinguish between international cooperative goods that are both partially excludable and partially rival, and those that are one but not the other. It helps us to tease out the distinct legitimacy effects of rivalness and excludability in international law that would otherwise not be as clear. It addresses the difference between excludability, which is often changeable by law, and rivalness, which is not. Specifying this difference in turn highlights the ability through exclusion of making public into club goods, and common pool resources into private goods.
Finally, using the language of public goods precisely facilitates our ability to see the different issue structures of different aspects of a broader international issue, and to design and implement different legal responses for those different aspects. Because using this language precisely means building identification of issue structure on analysis of underlying concepts such as rivalness and excludability, rather than on ad hoc categories such as best effort, aggregate effort, and weakest link, it allows a more nuanced reading of issue structure. It also makes clearer the links between structure (understood as some combination of rivalness and excludability) on the one hand and legal effect and potential legal solutions on the other. This helps us to understand the international legal ramifications of both the design of new mechanisms of international law, and litigation through existing mechanisms.
These various analytical observations help us to know which states need to be included in international legal mechanisms, and when nonstate actors should be involved as well. They show us when the focus of these mechanisms should be on deliberation, and when it should be on monitoring and enforcement. They suggest to us when international law can be soft, and when it needs to be hard. They help us to predict what elements of proposed or existing international cooperative law are most likely to trigger legitimacy concerns among participants.
More broadly, the narrower interpretation of public goods brings a set of useful analytical tools from economics to the study of international law. Leaving the analysis of rivalness and excludability to the economists to develop means that it will be developed in a way that addresses the concerns of economists, rather than those of students and practitioners of international law. That would mean little focus on issues such as legitimacy and enforcement; these issues are addressed in this article in a way not seen in the economics literature, and not possible in an international law literature that builds on a more generic understanding of public goods. Bringing this mode of analysis into the study of international law need not displace the deployment of more generic uses of the concept of public goods for the political purposes of developing a constituency for new efforts at international cooperation. But it is useful both to have the mode available when its analytical tools will be useful, and to keep clear the distinction between the more generic and more specific uses.
