Many IEPGs are strictly global. Examples include the conservation of the genetic diversity on which all future evolution depends, the mitigation of climate change, the control of emerging infectious diseases, and the management of sea areas beyond national jurisdiction. Many more are regional, such as the control of acid rain, the management of multi-country river basins, and the protection of international watersheds (Touza and Perrings 2011). Like all public goods, IEPGs exhibit both consumption indivisibilities and non-excludability. Non-excludability means that once the good is provided, none can be excluded from enjoying the benefits it confers. Indivisible consumption occurs when one country’s enjoyment of the benefits does not diminish the amount available for others. Public goods are said to be ‘pure’ when they are both non-exclusive and non-rival (indivisible) in consumption. They are said to be impure if they are either partially excludable or partially rival—the most common form of which are local public goods, particularly the local common pool resources analyzed by Ostrom (1990). In most cases, it is not possible for any single state to provide such goods on its own. International public good supply depends on either international coordination or international cooperation (Anand 2004).

This article focuses on IEPGs whose benefits extend to people in multiple countries. Such IEPGs frequently also deliver benefits across multiple generations (Kaul et al. 1999), but we do not address this aspect of the problem. In practice, the beneficiaries of international public goods include national populations and their representatives, nation states, transnational corporations and non-governmental organizations, as well as a newly emerging set of institutions. Globalization has altered the way that members of civil society organize themselves across national boundaries. The information revolution has also stimulated new forms of social participation. New networks, frequently built around environmental websites, enable the exchange of ideas and implementation techniques. These new relationships and interactions have created a ‘global environmental public’, interested in asserting new rights and responsibilities to the resources of the planet. Its concerns span both the ethical responsibilities of individuals, organizations, countries and corporations, and the alternative forms of governance of the biosphere.

Following the Millennium Ecosystem Assessment, we suppose that the benefits people obtain from biosphere depend on a set of ecosystem services comprising:

These services affect human wellbeing in many ways: through their role in the production of consumption goods, their support of human health and security, or the satisfaction of peoples’ cultural and spiritual needs. A number of these services have the characteristics of IEPGs, the most important of which involve the regulating and supporting services. Figure 1 indicates the relation between categories of ecosystem services and components of wellbeing identified by the Millennium Ecosystem Assessment. Of these, only the provisioning services consistently generate benefits that are both divisible (rival) and exclusive. The other services yield benefits that are generally indivisible and non-exclusive. We focus on the group of ecosystem services that are both public and international. These are services that: (i) cover more than one group of countries; (ii) benefit not only a broad spectrum of countries but also a broad spectrum of the global population; (iii) meet the needs of both present and future generations (Kaul et al. 1999; Anand 2004). International public goods generated in any one county must therefore generate spillover effects beyond a nation’s boundary (Morrissey et al. 2002).

IEPGs can further be classified according to their ‘technology of supply’ (Sandler 2004). The standard treatment of public goods focuses on demand (Hirshleifer 1983). However, understanding the technology of supply of IEPGs is critical to the development of appropriate incentives. Three common examples of public good supply technologies are ‘additive’, ‘best shot’, and ‘weakest link’ technologies. As the name implies, in the additive case, the socially available amount Y of a public good is nothing but the ‘simple sum’ of the separate amounts, yⁱ, produced by each of m participating countries, the i = 1,…, m. In the case of simple sum public goods, such as carbon sequestration, each unit of carbon sequestered has the same value no matter where it occurs. In the case of weighted sum public goods, such as habitat protection, the contribution of each hectare protected depends on its characteristics (Sandler 2004). For ‘best shot’ public goods, the benefit to all countries is determined by the most effective provider. For example, the Centers for Disease Control and Prevention are funded by the U.S.A., but provide information on infectious diseases to all countries. For ‘weakest link’ public goods, the benefits to all countries are limited to the benefits offered by the least effective provider. The best example of this is the control of infectious diseases. So for HIV and tuberculosis, the level of protection available to all countries is only as good as the control of the disease exercised in the poorest, most densely populated, and least well-coordinated country (Perrings et al. 2002).

Social composition functions

Of all the Millennium Ecosystem Assessment ecosystem services, the regulating services are most often supplied as IEPGs. Examples include disease control, which is frequently supplied as a weakest or weaker link public good, climate regulation through, e.g., carbon sequestration, which is supplied as an additive pure public good, or watershed protection which is generally an additive but impure public good (Holzinger 2001; Dombrowsky 2007; Touza and Perrings 2011). Many international public goods are also jointly produced with local public goods. Biodiversity in tropical forests, for example, yields a set of private benefits in the form of timber and other products including medicinal plants, hunting, fishing, recreation, and tourism. At the same time, tropical forests are a source of carbon sequestration, genetic information, hydrological and microclimatic regulation—commonly described as co-benefits (Perrings and Gadgil 2003).

An important feature of IEPG is that their spatial extent depends partly on the natural hydrological and atmospheric flows, and partly on the social linkages between countries—the flow of goods, people, and information. The global reach of carbon sequestration is a property of the general circulation system, but the global reach of disease regulation is a property of the global trade and air transportation systems. In fact, the closer integration of the world economic system has rapidly increased the number of environmental public goods that are global in reach (Kaul et al. 2003b):

Consider the conservation of endangered species. Can we rely on the national action to produce the efficient amount of an IEPG such as the protection of iconic species? The key to understanding this lies in the difference between a pure public good and a private good. For a private good, everyone pays the same price, but is free to consume as much or as little as they want. Consumers adjust the quantity they consume given the market price. For a pure public good everyone consumes the same amount of the “good” but is willing to pay a different price for it. Consumers adjust the amount they are willing to pay for the public good given the quantity supplied (Batina and Ihori 2005). In general, private provision of public goods will be below the socially optimal level. Efficiency requires that marginal benefit equals marginal cost. In the case of conservation of endangered species (or any other public good), the relevant measure of marginal benefits is social marginal benefit—the sum of all countries marginal benefits. We illustrate the problem in Fig. 2, in which local and global benefit curves for species conservation in a particular country, i, are presented. Global benefits are represented by the vertical sum of the benefit curves of country i, and all other countries.

The level of conservation in country i that maximizes local net benefits is indicated by yⁱ, while the level of conservation that maximizes global benefits is indicated by . yⁱ is given by the intersection of local supply and local benefit curves, and by the intersection of local supply and the vertical sum of local benefits for all countries. Since the marginal cost of provision at is greater than country i would be willing to accept on its own, socially optimal provision of the public good depends on the existence of a mechanism to cover the ‘incremental’ cost of socially optimal provision to country i.

Biodiversity conservation, like many other IEPGs, is an impure global public good. If there are many potential providers, each generates local benefits from its conservation effort, but also benefits from the conservation actions of others. Following, Perrings and Gadgil (2003), we characterize the problem for the individual country as follows. Vⁱ denotes welfare of the ith country, assumed to depend on consumption of a vector of market goods, xⁱ, and global biodiversity conservation, denoted C = C(y¹, y²,…,y^m), then the problem it faces is of the general form:

In other words, it derives a direct benefit from its own conservation efforts, yⁱ, but also benefits from the global conservation effort to which it has contributed, C. If the cost of conservation effort in terms of the cost of market goods is p(yⁱ), and if the income available to country i is Iⁱ, this will be subject to the constraint that

In the absence of cooperation, and noting that , wellbeing of the ith country will be maximized where the local level of conservation satisfies:

If global wellbeing is the sum of the welfare of all countries, , then global wellbeing will be maximized where

The extra terms in the summation term on the right hand side capture the conservation benefits that the ith country confers on all other countries. These benefits will be neglected by the ith country unless there is a mechanism to convert them into a direct incentive.

The failure of markets to signal the global benefit of such public goods accordingly results in under-investment in their local provision. The benefits of protection, management and establishment of forests provide a good example. Apart from the loss of the valuable environmental services (e.g., protection of genetic resources, air quality maintenance, climate regulation and regulation of human diseases), forest degradation frequently translates into a loss of timber and non-timber forest products important to local livelihoods (Landell-Mills & Porras 2002).

Currently, there are few measures of the underprovision of public goods. The United Nations Development Programme (UNDP) has opted instead for measures of “adequate” provision that differ from one public good to another. Such measures can, for example, correspond to the complete elimination of global public bads. More generally, they are measures of what is considered possible, given the current state of technology (e.g., to control—rather than eradicate—the problem of HIV/AIDS) and what is “fair” (e.g., what would emerge if all concerned stakeholders had an effective voice in the decision-making process) (ODS-UNC 2002). The criterion of adequacy is not meant to indicate optimality—the balancing of marginal costs with the sum of people’s marginal willingness to pay for a particular public good (see Samuelson 1954; Cornes and Sandler 1996). Rather, it is meant to establish a relatively simple, yet reliable, yardstick for measuring the present provision of a certain good against a technical notion of adequacy.

Systems of Payments for Ecosystem Services (PES) have become popular instruments for dealing with IEPGs, in part because they appear to satisfy the incremental cost principle (Ferraro and Simpson 2002; Goldstein et al. 2006; Wunder 2007; Ferraro and Kiss 2007; Pagiola 2008; Engel et al. 2008; Wunder et al. 2008). They are not, however, appropriate mechanisms in all cases. International PES schemes are appropriate where non-marketed ecosystem services are privately supplied in one country, but offer benefits that are public and accrue elsewhere.

To illustrate the potential pluses and minuses of PES, we consider a particular problem: the impact of local deforestation on the provision of a range of IEPGs including climate regulation (through carbon sequestration), protection of genetic diversity, and watershed protection in addition to timber and non-timber forest products. The Economics of Ecosystems and Biodiversity (TEEB) study is a major international initiative to draw attention to the global economic benefits of biodiversity, to highlight the growing costs of biodiversity loss and ecosystem degradation, and to draw together expertise from the fields of science, economics and policy to enable practical actions moving forward. The current assessment of TEEB on PES has used existing studies to estimate the mean value of both the macroclimatic regulation offered by terrestrial carbon sequestration, and the change in provisioning and cultural services offered by forest systems. Its findings are preliminary but telling. TEEB (TEEB 2009; Kumar 2010) suggests that the mean values of forest ecosystem services, in US$/ha/year, are dominated by regulatory functions: specifically regulation of climate ($1965), water flows ($1360), and soil erosion ($694). The mean value of all provisioning services combined—timber and non-timber forest products, food, genetic information, pharmaceuticals—is $1313. This is less than the value of water flow regulation alone. There are substantial off-site benefits to forest conservation that are not currently captured by forest landowners and are difficult to incorporate on PES schemes.

Governments around the world have frequently implemented forest protection policies in areas high in biodiversity, landscape beauty or critical for their watershed protection. However, as pressure mounts on governments to curtail spending and cut budget deficits, their ability to invest directly in the provision of public goods and services is compromised. Where public authorities have been unable to tackle the public good problem, they have searched for ways to involve non-governmental actors. Efforts to transfer responsibility for forest environmental services out of the public sector have relied on a combination of regulation and market-based approaches (Landell-Mills and Porras 2002). Experience has shown that well-designed market-based instruments can achieve environmental goals at less cost than conventional “command and control” approaches, while creating positive incentives for continual innovation and improvement (Stavins 2003). Examples of such instruments in the forestry sector include stumpage value-based forest revenue systems, financial and material incentives, long-term forestry concessions, trade liberalization, forest certification and the promotion of markets for non-timber forest products.

The costs and benefits associated with many human activities spill over jurisdictional boundaries, thereby generating externalities that are often reciprocal and quantitatively significant (Cornes 2008). Therefore, IEPGs supply depends on either international coordination or international cooperation. Among payment schemes to internalize the external benefits of maintaining intact forests, Reducing Emissions from Deforestation and Forest Degradation (REDD) is an effort to create a financial value for the carbon stored in forests, offering incentives for developing countries to reduce emissions from forested lands and invest in low-carbon paths to sustainable development. REDD is an example of international coordination in delivery of ecosystem services. Its integration into international market-based climate change policies poses a number of challenges both to institutional design and to implementation. At present, for example, there are few effective mechanisms for converting international payments to governments into incentives to on-the-ground forest communities (Myers 2008; Sikor et al. 2010). Indeed, creating an effective multilevel system of payments is seen as the core issue in building REDD considering that REDD goes beyond deforestation and forest degradation, and includes the role of conservation, sustainable management of forest and enhancement of forest carbon stocks (Anngelsen and Wertz-Kanounnikoff 2008). It is predicted that financial flows for greenhouse gas emission reductions from REDD could reach up to US$30 billion a year. This significant North–South flow of funds could reward a meaningful reduction of carbon emissions and could also support new, pro-poor development, help conserve biodiversity, and secure vital ecosystem services.

Public goods are recognized as having benefits that cannot easily be confined to a single “buyer” (or set of “buyers”). Yet once they are provided, many can enjoy them for free. A clean environment is an example. Without a mechanism for collective action, these goods will generally be underprovided. In fact, many crises dominating the international policy agenda today reflect the under-provision of global public goods. With globalization, externalities are increasingly borne by people in other countries. Indeed, issues that have traditionally been merely national are now global because of the greater interconnectedness of the planet.

Kaul et al. (2003b) suggest a rethinking of three notions underpinning the theory of public goods. First, properties of non-rivalry in consumption and non-excludability of current benefits do not automatically determine whether a good is public or private. Some goods may be either public or private. Nevertheless, it is important to distinguish between a good’s having the potential of being public (that is, its having non-rival and non-excludable properties) and its being de facto public (non-exclusive and available for all to consume). Second, public goods do not necessarily have to be provided by the state. Many other actors can, and increasingly do, contribute to their provision. And third, a growing number of public goods are no longer national in scope, having assumed cross-border dimensions. Many have become global and require international cooperation to be adequately provided.

For the most part, the theoretical and empirical literature in economics has focused on two polar models of public goods provision: the provision of pure public goods that benefit all agents, and the provision of local public goods that only benefit agents in one community (Bloch and Zenginobuz 2007). We are concerned with cases where the members of one community enjoy positive spillovers from the public goods provided by other communities. In the context of global climate regulation, the REDD scheme will compensate tropical nations that succeed in reducing carbon emissions from deforestation and forest degradation—source of nearly one fifth of global carbon emissions. Since forests offer a number of benefits aside from carbon, however, the scheme could potentially benefits to communities that would otherwise be unable to afford them (Stickler et al. 2009). If well designed and implemented, PES schemes such as REDD have the potential to secure provision of IEPGs that offer benefits at multiple scales, such as the protection of water supplies, local and regional climate regulation, and habitat provision for the protection of biodiversity. The effectiveness of PES schemes depends heavily on the conditionality of payments (Arriagada and Perrings 2009), but the principles for their effective design and implementation are well understood.

Rodrigo Arriagada, Email: rarriagadac/at/uc.cl.

Charles Perrings, Email: Charles.perrings/at/asu.edu.