Historically, governments have adopted legalized, integrated, and global rules to govern oligopolistic industries, such as shipping, chemicals, and industrial production. By contrast, they have adopted nonbinding and unintegrated rules and institutions to govern competitive industries, such as energy, agriculture, and mining, at the national or subnational scale. Considering that competitive producers face greater barriers to political collective action, what explains the form of global governance across these sectors? This article demonstrates that oligopolistic producers are more intensively and extensively regulated than competitive markets because producers in oligopolistic industries can more cost-effectively alter markets to meet environmental goals. Therefore, despite their political influence, oligopolies are regularly called upon to initiate and sustain market transformation on a global scale. New qualitative evidence from two treaty regimes governing different types of markets supports this theory, as well as new quantitative data on the full range of global environmental treaty regimes since World War II.
During the final hours of negotiations on the Minamata Convention on Mercury, negotiators were settling the contentious issues of financing and implementation. Throughout the negotiations, they maintained that they could not have the same expectations across the industries and sectors that contribute to atmospheric mercury pollution. In particular, the oligopolistic industrial sectors would have legally binding and precise deadlines to meet. By contrast, negotiators decided to treat the main economic source of mercury pollution in a wholly different way than the capital-intensive industries, by providing for imprecise and nonbinding rules to reduce mercury use in artisanal gold mining.
The variable forms of governance adopted under the new convention for mining and industry are not unlike what governments have adopted under other global environmental treaties governing multiple markets. Nor is it unlike the much wider variation in the forms of governance across global environmental regimes. In fact, governments have historically adopted more stringent rules to regulate oligopolistic industries than competitive industries in protecting the global environment. Despite their political influence, oligopolistic industries have encountered more integrated, legalized, and globalized rules than competitive industries. What explains this pattern?
Consider two important aspects of global environmental governance. First, governments have adopted legally binding rules to govern shipping, chemicals, and industrial manufacturing. However, they have adopted nonbinding rules more frequently to govern competitive markets for agricultural goods, forestry products, and energy production or consumption. Second, governments have tended to integrate new rules for the industrial and shipping sectors, but not for the agricultural and land-intensive ones over time. These patterns of global environmental governance have become more pronounced. Regimes consisting of integrated rules and institutions impose legal obligations on industries on a global scale (allowing for national-level implementation). By contrast, regimes consisting of unintegrated rules and institutions tend to impose a combination of binding and nonbinding rules on industries, relying on national and subnational efforts to tailor guidelines to fit specific local circumstances.
This pattern does not match prevailing views on the political economy of regulation and public policy. We would generally expect that powerful corporations would block costly new rules that would alter market conditions, unless those changes would benefit incumbent businesses. Otherwise, they would seek to influence new regulations to solidify or enhance their market positions. Generally speaking, these oligopolistic corporations and their lobby groups would oppose new environmental controls because new controls would raise compliance costs. If unwanted regulations become inevitable, however, they would hedge against the new rules to minimize the expected compliance costs and maintain their market positions (Meckling 2015).
The landscape of global environmental governance suggests that this prevailing viewpoint is only partly correct. It does not account for the relatively stringent regulation of oligopolistic industries, which runs counter to the expectation that markets with many small-scale producers will face more demanding expectations to limit their environmental pollution, because of the barriers to collective action by large groups. In this respect, our understanding of the institutional form of global environmental governance is incomplete because of the limited capacity of “regulatory capture” theory, Olsonian collective theory, and neo-pluralist theory to explain the institutional design of environmental regimes. These perspectives on the political economy of institutional design focus more on the agency and capacity of oligopolistic industries to influence policy than they do on the relevance of those market structures in policy implementation—or on the constraints on global institutional design stemming from various market structures.
Market structures are important not only because they limit or enable market actors to influence policy-makers, but also because they have implications for the implementation of environmental rules in those markets. In particular, oligopolistic markets possess two characteristics that make them more cost-effective than competitive markets in implementing new rules on national and global scales. First, technical innovation and diffusion of new technologies are more likely in oligopolies than in competitive markets, because of disincentives to invest in risky innovations and the limited resources of producers in competitive markets. Second, oligopolies have a reduced need for differentiating and tailoring technological or process-oriented changes when implementing new regulations in their markets, and the number of firms (at upstream levels of the market) in oligopolies limits the number of key businesses who would need to make technical and informational changes so as to transform the downstream market. Innovation prospects and industrial organization jointly enable oligopolistic industries to meet new environmental standards more cost-effectively than competitive markets. By contrast, governments would need to make considerable public resource investments to bring about practical improvements in competitive markets.
In designing global regimes, governments utilize markets to the extent they can. They seek to design the regime to fit the market by harnessing the capacities and resources of producers to adopt new technologies or practices with better environmental properties. However, small-scale producers in competitive markets generally have very limited resources and market power, which means that adopting legally binding targets and timetables invites a higher risk of noncompliance than in governing oligopolistic industries. Moreover, since oligopolies can more readily initiate and sustain market transformations on national and even global scales than can producers in competitive markets, governments are more inclined to adopt globalized rules to govern oligopolies, which can later be modified or amended in the context of the existing regime as new regulatory demands arise. The capacities of producers and market structures, therefore, influence government decisions over the legalization, standardization, and integration of environmental rules.
This article evaluates a market structures explanation for the design of rules and institutions in global environmental regimes, on the basis of original qualitative and quantitative evidence. The regimes governing mercury pollution and ozone-depleting substances illustrate that governments seek to regulate markets differently, even within the same regime, because of basic market characteristics. Under the mercury treaty, governments adopted “bottom-up,” nationally centered, and nonbinding expectations for artisanal gold mining, but did largely the opposite for most industrial sources of mercury pollution. Although the institutions helping prevent atmospheric ozone depletion by manufactured substances are an example of integrated and legalized global governance, the use of methyl bromide in agriculture has been handled differently than the use of manufactured industrial chemicals under the Montreal Protocol. Market incentives and competition have enabled more technical changes and implementation in the oligopolistic industries, leading to fewer exemptions from legally binding deadlines than in the agricultural applications.
Furthermore, an analysis of the agreement histories and development of nineteen global environmental treaty regimes over six decades demonstrates that industrial sectors (e.g., shipping, chemicals) have been subject to more legalized, globalized, and integrated regulation than competitive industries (e.g., energy, agriculture, forestry). New quantitative data on the participants in meetings of these global regimes also indicate that the regimes with more integrated, globalized, and legalized regulations have the highest concentrations of business representation at their meetings, despite the general opposition of big business to new global environmental controls and the strategy of hedging against new global agreements that appear inevitable. In that respect, the regimes whose meetings are dominated by business associations impose intensive and extensive rules on powerful corporations.
Recent research on the institutional design of international regimes has highlighted the importance of political economy (e.g., Mattli and Woods 2009). The market structures explanation builds on this research, although its predictions differ from the existing perspectives. Three political economy theories stand out: Olsonian collective action theory, “regulatory capture” theory, and neo-pluralist theory. These theories favor a nuanced set of expectations for institutional design, providing a more microfoundational perspective than structural theories (e.g., Krasner 1991).
A starting point for a political economy perspective of international regimes is Olsonian collective action theory, according to which stakeholders have varying degrees of influence over regulatory choices, based on their group size. Large stakeholder groups suffer from more severe free-rider problems hampering collective action than do small groups, and they do not supply public goods for themselves as extensively as small groups do for themselves (Olson 1965). Consistent with this expectation, internationalized trade associations represent oligopolistic industries in multilateral negotiations on climate change and biodiversity (Orsini 2011). Civil society is not nearly as organized as multinational corporations, despite the growing number of civil society organizations engaged in some environmental negotiations (Hanegraaff 2015). Interest representation at global environmental meetings supports a group size theory of collective action and policy influence.
Second, the design and substance of global rules and regulations seem to indicate that organized interest groups have “captured” some international regimes nominally designed to provide for the public interest (Mattli and Woods 2009). Studies of ozone-depleting substances have made this claim regarding industrial and agricultural chemicals (Oye and Maxwell 1994). To the extent that new regulations favor the market interests of multinational corporations, it is common to find evidence that corporate lobbying played an important role in the specific provisions of the international settlement, and especially in its implementation (Urpelainen 2011; Marcoux and Urpelainen 2014).
Third, a neo-pluralist perspective on policy design emphasizes cleavages among societal interests on specific issues (Dahl 1961). According to neo-pluralism, the business community is too heterogeneous to have consistent and complementary interests in international politics, making it difficult for cohesion to last long and for a singular business perspective to emerge in national or international decision-making processes (Falkner 2009). Neo-pluralism provides some insight into why international environmental governance does not consistently or overwhelmingly protect the interests of multinational corporations, since new regulations have been adopted against the preferences of major lobby groups (Orsini 2011).
Although these political economy perspectives are able to explain certain patterns of environmental regime design, they do not account for the ways in which governments have chosen to regulate different industries to mitigate their pollution. In particular, oligopolies are more stringently regulated than we would expect, especially since they typically resist new regulations before they become inevitable, while competitive industries are not intensively regulated, and yet cannot lobby effectively at the global level to prevent stringent regulations. Olsonian collective action theory and regulatory capture theory would predict that oligopolies should be able to avoid new regulations, lower compliance costs considerably, or generate private benefits. These theories do not explain the legalization, integration, or standardization of global rules to govern oligopolistic industries, or the lack thereof in governing competitive industries. Neo-pluralism also cannot account for this pattern, even though it does temper expectations that large corporations will be able to control new global regulations.
The next section specifies the conditions under which governments develop integrated, standardized, and legalized global regulations protecting natural resources, by elaborating on the constraints imposed by market structures in designing environmental regimes.
Regime Design Under Market Constraints
Markets not only condition the scale and effectiveness of business lobbying, but also the constraints on implementing international environmental regulations. I consider two ideal types when referring to market structures: oligopolies and competitive markets. These markets differ according to the number of incumbent businesses in the market as producers. Specifically, oligopolies have considerably fewer producers than competitive markets, although there is no single numerical threshold.1
In particular, oligopolistic businesses tend to have the resources and market power to reduce environmental pollution more cost-effectively than do businesses in competitive markets. Despite the capacity of oligopolies to block new regulations or influence their design, the same characteristics enabling them to exercise influence over new environmental controls also make them attractive instruments of environmental change. Governments can rely on oligopolies to carry out technical and practical changes that transform national or global markets, particularly when the main producers are organized transnationally and have access to foreign markets. In this respect, governments seek to provide oligopolies with new incentives to transform practices or technologies in the markets they dominate by crafting regulations accordingly.
By contrast, competitive markets tend not to have the unified lobbying campaigns of oligopolies. Individual producers will rarely have clout in political circles, unlike the producers in oligopolistic industries (Orsini 2011). Ironically, this will also mean that governments are less inclined to expect producers in competitive markets to make substantial changes in their routine practices or technologies, since market pressures and overwhelming resource constraints would disincentivize producers to make such investments. Competitive markets will be more resource-intensive than oligopolistic ones, in part because the level of competition incentivizes market participants to compete by producing more, which generally requires higher resource inputs (Lambertini 2013).
Generally, governments would prefer to economize on the transaction costs of implementing environmental measures, since public resources are limited and governments are reluctant to make such public-resource investments when the benefits to their publics are uncertain or the resources would need to be redistributed among a large number of countries. Governments place value on harnessing markets to provide public goods, to conserve public resources (Andonova 2010), yet the constraints on market participants condition what governments can expect of them.
In regulating oligopolistic industries, governments tend to form integrated rules and institutions, to economize on transaction costs by avoiding duplication or unnecessary overlap among independent agreements. Instead, governments tend to form integrated regimes consisting of agreements that establish new rules and institutions building on earlier agreements. This enables governments to harness the knowledge and institutions already in place to manage new challenges under the mandate of a founding agreement. Governments are especially inclined to form integrated regimes to govern oligopolies because, as new concerns arise, they can turn to the same oligopolistic producers to implement a new set of changes and diffuse those changes to downstream segments of their markets. The market power of oligopolistic companies and the stability of those markets inhibit the need for subglobal or independent institutions. An integrated regime can economize on transaction costs by providing for standardized obligations among all market participants in the industries under the regime’s mandate.
Developing integrated regimes to govern competitive markets is far less productive, because these markets are less responsive to the incentives that globalized rules can provide them. The disincentives to make environmentally friendly changes in agricultural markets, for example, remain in both wealthy and less wealthy countries alike. Price competition in energy markets has actually prompted investments in fossil fuel extraction that further lower the price of carbon-polluting energy sources (e.g., natural gas from shale deposits). Under these circumstances, a single set of incentives created through global negotiations will not generate the market responses that they would in oligopolistic industries, in which well-organized producers can decide how best to change practices to comply with the new regulations without severely hampering their market shares and revenues. Rather, governments will tend to form unintegrated rules and institutions to provide incentives and resources to specific market stakeholders or in specific local areas when governing competitive markets.
Much like the decision to integrate rules and institutions, governments also decide on whether to legalize obligations on the basis of market structures. In legalizing commitments, they consider the potential for legal noncompliance. They also are subject to the lobbying efforts of businesses on the key issue of legalization. Oligopolies tend to prefer legalizing commitments, if they view regulations as inevitable, to avoid artificial competitive disadvantages. National producers who face domestic environmental controls have lobbied their governments to “internationalize” those controls to reduce the competitive disadvantages resulting from regulatory differences (DeSombre 2000). Similarly, governments also tend to prefer legalizing commitments when governing oligopolies, not only because domestic businesses would prefer to avoid uneven regulations that disfavor them, but also because legally binding commitments are more informative than nonbinding ones. Binding rules provide more information about what other countries’ businesses will do (Haas et al. 1993).
However, legalizing commitments when governing competitive markets would entail either broad and imprecise legal provisions or a higher probability of noncompliance if the provisions are made specific. Although broad legal mandates can prove useful, particularly in setting out a mandate for future governance, imprecise provisions can cause uncertainty and disputes, as well as few substantive efforts to comply. On the other hand, specific legal commitments made to govern competitive industries can raise the prospect of noncompliance, because of the constraints on market transformation in these industries. Perhaps the most notable example comes from the Kyoto Protocol, under which some wealthy countries either overshot their emissions targets or purchased enough carbon credits to satisfy their legally binding commitments.
Two expectations follow from this argument. First, governments will more likely form integrated regimes with standardized obligations when governing oligopolies than when governing competitive industries. Second, governments will more likely make legally binding agreements when governing oligopolies than when governing competitive industries.
Hybrid Governance of Industries Within Regimes
I evaluate the importance of market structures in environmental regime design by analyzing how governments have regulated oligopolistic and competitive industries within a single regime. This empirical strategy eliminates the possibility that issue-area differences would explain different regime designs, since the issue-area is the same within a regime, but the industry structures vary. Moreover, a focused comparison of institutional responses within a regime eliminates other potential threats to inference, such as variation in the governments involved in the regime or in its timeframe. Therefore, such an analysis can identify the mechanisms driving different institutional designs. I consider two global regimes that each regulate both oligopolistic and competitive industries, to provide a microlevel assessment of the market structures argument: the ozone layer regime and the mercury regime.
The Montreal Protocol on Substances that Deplete the Ozone Layer has been incredibly effective in mitigating the risks of stratospheric ozone depletion and lowering the emission of climate-warming substances. Governments have successfully phased out entire classes of industrial chemicals under the Protocol and a series of amendments that were adopted during the 1990s, with the developing countries following suit with a ten-year lag period and financial assistance. However, governments have not phased out an agricultural pesticide known as methyl bromide. Several of the wealthiest countries in the world have continued to use methyl bromide, well past the 2005 deadline for a complete phase-out, after requesting and receiving exemptions under the Protocol. The countries that led efforts to reduce the use of ozone-depleting substances (ODSs) in industrial and manufacturing applications have been responsible for the continued use of agricultural chemicals that deplete stratospheric ozone. This divergence stems from the markets in which these chemicals are used.
The markets in which industrial ODSs (and their replacements) are used are different from the markets in which methyl bromide is used. Two characteristics distinguish those markets: the number of intermediate producers who rely on the chemicals in their retail items, and the price sensitivity and product substitution opportunities of consumers of those retail items. At the intermediate and consumer levels, the markets involving industrial ODSs have been more conducive to technical innovation and transnational technological diffusion than have the agricultural markets consuming methyl bromide.
According to a long-time DuPont representative who attended the Montreal Protocol meetings, DuPont does not invest in replacements for methyl bromide as much as it has in industrial chemicals because the market for agricultural substitutes is “too fragmented.”2 In other words, the return on investment is far less risky in the industrial sectors, especially air conditioning and refrigeration, where those chemicals are widely used as coolants. The fragmented markets for methyl bromide do not incentivize capital investment in substitutes because consumers of the substitutes—namely, farmers—are highly competitive, with thousands of small-scale producers selling to markets in which comparable substitutes are available.
The main use of methyl bromide in recent years has been in strawberry production, especially in the state of California. Consumption of strawberries is price-sensitive, and substitutes are available. Although the markets for industrial chemicals with ozone-depleting potential are also competitive, they are much more oligopolistic, and the risks and market pressures on producers do not generate the disincentives to innovate or adopt new technologies to the same extent as in the agricultural markets employing methyl bromide.
Consequently, governments have handled these two sectors differently under the Montreal Protocol, despite the facts that both industrial and agricultural ODSs are under the mandate of the Protocol and that the parties, institutional rules, and timeframes corresponding to both sets of chemicals are identical. Figure 1 shows that the wealthy non-Article 5 (non-A5) states under the Protocol have nearly entirely phased out the consumption of ODSs. However, the share of remaining ODS consumption represented by methyl bromide increased during the period when industrial ODSs were being phased out in the wealthy (non-A5) states. Even though the consumption of methyl bromide has also declined in absolute terms, albeit not year on year, enough consumption remained in the non-A5 countries throughout the 1990s and 2000s to make it 50 percent of their total ODS consumption (in metric tons) as recently as 2012.
United States businesses have been much more in favor of new regulations on industrial than on agricultural ODSs. Lobbying efforts by American businesses have sought to promote changes in the industrial sectors but stasis in the agricultural sectors under the Montreal Protocol, reflecting the downstream economic incentives of US chemical producers and strawberry farmers in California. Consequently, regulatory developments and rule implementation across these sectors have diverged over time. Although it is under strictly the same legal controls as industrial ODSs, methyl bromide has been the subject of more exemption requests than any other single ODS, perverting the original intent behind the so-called “critical-use exemptions” (Gareau 2015).
Mercury Use Mitigation
Recently, governments have adopted a similar strategy to reduce atmospheric mercury pollution. With the 2013 Minamata Convention on Mercury, the international community completed a treaty to control mercury emissions from over a dozen industries worldwide. This required governments to design a treaty that would anticipate the challenges of mercury mitigation in a wide variety of industries with different market structures. They responded with a treaty reflecting both the lobbying efforts of many businesses and the constraints implied by their market structures.
The sector most singled out under the mercury treaty is the largest contributor to atmospheric mercury emissions, but it is also widely believed to present the largest challenge to the international community in reducing human mercury exposure. Namely, artisanal and small-scale gold mining (ASGM) was named the sector making the largest contribution to atmospheric mercury emissions (37 percent of the total) soon before the treaty was negotiated in January 2013. Just as importantly, ASGM had become the largest sectoral source of mercury pollution in a relatively brief timespan: Earlier, in the 2008 Global Mercury Assessment, it was named the second largest source, with just over 18 percent of the global emissions total.
Unlike the industrial sectors using mercury, the gold market is globalized and highly competitive. There is a single world gold price, with the supply coming from industrial mining companies and uncertified illegal miners. Although industrial mining companies use mercury-free processing to extract gold from ore, millions of artisanal miners use mercury because of its simplicity to recover gold. Artisanal miners are located in as many as 70 countries, and the largest supplies of artisanal gold come from West Africa, Southeast Asia, and the Andean region. Due to the social and economic conditions and resources of artisanal miners, they lack institutionalized organizations or networks that span multiple countries.3 They also have very limited access to capital for purchasing mercury-free technology and lack the technical expertise to use those technologies.
By contrast, industrial producers employing mercury, or whose processes emit mercury as a byproduct, are generally oligopolistic. For example, the chlorine industry is organized nationally, transnationally, and globally into trade associations. Most chlorine production has historically involved mercury, until recent movements by the industry toward mercury-free processes. The capital resources and oligopolistic structure of the producers has enabled them not only to lobby reasonably effectively for their interests, but also to diffuse information about the technical processes and capital expenditures needed for conversions.4 Moreover, production conditions and circumstances for chlorine are much less varied than in artisanal gold mining, which often occurs in remote areas and impoverished communities with social and cultural conditions that hamper standardized solutions.5
Under the Minamata Convention, countries with ASGM are not required to meet a legally binding phase-out deadline. The treaty does not stipulate standardized expectations for those countries, except for some minimal requirements for filing so-called National Action Plans to reduce and, where possible, eliminate mercury use in ASGM. However, those plans are voluntary: A country can choose not to submit a plan, which would prohibit that country from receiving funds to mitigate mercury use in ASGM under the financial mechanisms of the treaty. ASGM requirements on countries with those sectors are minimal, and ultimately the treaty leaves it to those countries to decide for themselves whether to exercise the provisions needed for assistance.
The obligations on nearly all industrial sectors covered by the treaty are legalized, global, and carry specific deadlines. The use of mercury in chlorine production is expected to end in all member states by 2025. The use of mercury in producing lamps, cosmetics, pesticides, and measuring devices is expected to end in all member states by 2020. Well-organized industrial sectors lobbied heavily against the treaty in its original form, which was intended to regulate all heavy metals (mercury, lead, and cadmium), but they acquiesced during the negotiations at the UN Environment Programme’s Governing Council and accepted a narrowly defined treaty on mercury alone. This “opposition first, hedging second” strategy has recently been theorized (Meckling 2015).
The obligations on the industrial and manufacturing sectors are the same across countries, unlike the country-specific approach adopted for ASGM. Since the National Action Plans for ASGM are expected to vary across member states, governments have constructed a “subregime” under the Convention that encourages country-specific or subnational solutions for reducing mercury use in the artisanal gold mining sector. Although future amendments to the Convention could be made to include more oligopolistic industries emitting mercury (e.g., petroleum production), no future agreements are envisioned that would shift the focus away from the country-specific voluntary approach to regulating the ASGM sector. Governments adopted a “top-down” approach to reducing mercury emissions from the industrial sectors, much like in the Montreal Protocol’s governance of ODSs, but simultaneously created the framework for “bottom-up” country-specific solutions in handling artisanal gold mining at community levels.
Generalizing the Argument
The ozone and mercury regimes illustrate that oligopolies provide more opportunities for governments to design rules targeting industries at the global level with legalized and standardized obligations. By contrast, competitive markets force governments to handle individual or localized cases separately. Competitive pressures do not enable producers in these markets to respond to globally set standards, or requirements and the resource constraints of producers inhibit their capacity to diffuse technology and information across long distances or different settings. Consequently, governments design global regulations differently, depending on the constraints imposed by the markets they seek to govern.
To generalize the argument, I collected original data on all global environmental treaty regimes since World War II. The data provide a historical and comparative perspective on how states have governed environmental issue-areas. The macro-level data complement the focused comparison of within-regime market regulation in the mercury and ozone regimes.
I investigated how markets shape the development of global environmental governance by analyzing two original datasets. One is based on the global regimes (or “treaty processes”) governing nineteen environmental issue-areas, and the second measures business representation in various global environmental meetings. The regime data identify patterns of legalization, standardization, and integration in global governance across a wide range of environmental issue-areas, while the business data identify patterns in the political representation of producers across regimes.
A treaty process governs each issue-area. Founding treaties initiate the process, and may lead to further treaties or nonbinding agreements that create a system of regulation. A treaty process is defined by four criteria: (1) it was established by a founding convention, (2) the founding convention is a legally independent agreement and not subsidiary to another agreement, (3) the founding convention covers a global problem, and (4) the founding convention includes or provides for meetings of parties to the convention. The sample of treaty processes came from all agreements satisfying these four criteria.
The sample consists of nineteen treaty processes that were measured yearly through 2014 (N = 602), relying mainly on the International Environmental Agreements Database Project (Mitchell 2002–2015). The founding conventions were negotiated in different years, all after 1945, and nine were negotiated since 1992. Table A1 lists the treaty process subjects, founding conventions, and start dates, and provides information on the agreement text sources (http://www.mitpressjournals.org/doi/abs/10.1162/GLEP_a_00368).
I measured the integration, legalization, and globalization of regimes with international agreements by using keywords (the online appendix provides details). In particular, integrated regulation takes the form of international agreements that build on one another to manage a specific issue that had been regulated under an earlier agreement. The sequence of agreements creates a system of regulations that fit together. The foremost instance of this occurs when governments adopt an amendment or protocol; a protocol or amendment is legally and operationally dependent on an earlier agreement.
By contrast, unintegrated regulation takes the form of international agreements that have similar or overlapping goals and subject matters but are not operationally or legally dependent on one another. They may share goals and mandates, but no legal or institutional links connect them. For example, efforts to reduce tropical deforestation have prompted governments to issue multiple political declarations on tropical forestry, including the 1985 Tropical Forestry Action Plan. These documents were issued outside the purview of the International Tropical Timber Organization (ITTO) and did not reflect regulations under the ITTO, which was established to manage the international trade in tropical timber.
Some integrated agreements are treaties, and others are nonbinding instruments. Nonbinding instruments were included because the market structures argument proposes when we should expect legal treaties and when nonbinding instruments. Some unintegrated agreements are also treaties, and others are nonbinding agreements. According to this measurement, the global governance of an environmental issue-area becomes more unintegrated with more such agreements year after year.
The treaty processes govern a wide range of industries. Consistent with the collective action literature, trade group representation in the treaty processes reflects the economic concentrations of producers. Oligopolistic producers have trade associations at national and international levels, but more competitive markets with many producers lack political representation at those levels, at least in the context of meetings of the treaty processes. I used the trade group data to analyze the impacts of oligopolistic and competitive industries on the legalization, globalization, and integration of global environmental governance.
Specifically, I analyzed the industrial characteristics of producers on a global scale, since the issues are defined as global because of the transboundary nature of externalities and the avowed need for international collective action. The key was to identify the primary markets governed under the treaty processes and to standardize the measure of market concentration across diverse issue-areas.
The trade groups sample was compiled from the Lists of Participants (LOPs) from meetings of the treaty processes. Nearly all of the processes had at least one official meeting, except for the one governing international rivers. Some have had over twenty official meetings. In fact, the rationale for referring to them as “processes” is to highlight that they are intended as negotiations that occur through successive meetings, with no termination envisioned. Nongovernmental representatives at the meetings come from businesses, the environmental community, academia, and international organizations. Collectively, the LOPs cover sixteen treaty processes and 174 meetings.6
The participants at the meetings provide an accurate representation of the industries with a stake in a treaty, because otherwise participants from those industries would not attend the meetings. Microsoft Inc. representatives do not attend meetings of the Minamata Convention on Mercury, but representatives from the European cement industry do. This is because Microsoft has no direct stake in the treaty, but cement producers do, because their operations may be affected by what governments decide under the Convention.
The LOPs indicate which industries have the greatest stake in a treaty regime, and their political organization. I drew inferences on the industrial concentration of producers on the basis of their trade groups at the international meetings, assuming that their industrial organization shapes their capacities to overcome barriers to political collective action. This follows a long history of research on group size, economic organization, and political organization spanning many policy fields (Olson 1965; Drope and Hansen 2009; Gawande and Magee 2012).
Since nearly all of the treaty processes have had meetings with nongovernment participants in attendance, this provides a standardized measure of business stakeholders across the treaty processes governing diverse environmental issue-areas. This measure has been used recently to analyze the relative abundance of business and environmental groups at the UN climate change meetings (Hanegraaff 2015). I focus on trade associations at global meetings to compare business dominance across different treaty processes.
To supplement the LOPs, I used industry reports published by the market research company IBISWorld to compare the market concentrations of specific industries. These reports provide independent data on the economic and producer characteristics of many industries represented in the treaty process meetings.
The Market-Constrained Development of Global Regimes
The majority of treaty processes with the highest ratios of trade groups to all nongovernmental groups at the meetings govern industrial and chemical sectors. These treaty processes also have among the fewest total participants at their meetings; there is a −0.21 correlation between total meeting participants and the ratio of trade groups to all nongovernmental groups (p value < 0.05). This negative correlation is consistent with the finding from a recent study of nongovernmental groups at the climate change meetings (Hanegraaff 2015). On the basis of this negative relationship, producer groups do not dominate larger meetings as much as they do smaller meetings. Industry groups are most prevalent in meetings of treaty regimes that draw less political or popular attention (Urpelainen 2011).
In particular, the following treaty processes have the highest mean ratios of industry-related to agriculture- or land-related participants (mean ratios are in parentheses): International Maritime Organization (0.959), Minamata Convention (0.813), Basel Convention (0.715), Montreal Protocol (0.738), and Stockholm Convention (0.671). All these treaties regulate pollution from chemicals, shipping, or industrial waste. According to IBISWorld Industry Reports, these sectors often have concentrated producers nationally, and in some cases globally. For example, the International Maritime Organization regulates shipping and ship-manufacturing producers that are politically organized at the global and continental scales. The four largest shipping companies had 44 percent of total shipping tonnage in 2014, up from 32 percent in 2010, and 34 percent of total shipping units. The air conditioning industry has comparable levels of industrial concentration at upstream levels of a national market, which facilitated conversions under the Montreal Protocol.
By contrast, the majority of treaty processes with the lowest ratios of trade groups to all nongovernmental groups at the meetings govern agricultural and forestry sectors. These treaty processes have among the most total participants at their meetings. The following treaty processes have the lowest mean ratios of industry-related to agriculture- or land-related participants (mean ratios again in parentheses): UN Forum on Forests (0.316), International Treaty on Plant Genetic Resources for Food and Agriculture (0.204), Convention on Biological Diversity (0.111), International Tropical Timber Organization (0.295), and Convention to Combat Desertification (0.16). Except for the Convention on New Varieties of Plants, which has limited state membership and is fairly esoteric, these are among the lowest trade association ratios of all the treaty processes.
According to IBISWorld Industry Reports, these sectors have been highly competitive with many producers in national and world markets. In agriculture, family farms account for 90 percent of all farms in the United States and produce 80 percent of world food in value terms (FAO 2014). In forestry, logging companies generally locate operations subnationally, relying on supply chain operations within a certain kilometer radius. In fact, deforestation in Indonesia has been shown to increase as the territories of forest managers decrease, reflecting a Cournot competition model in which more oligopolistic control over natural resources helps limit environmental degradation because oligopolistic producers withhold supply (Burgess et al. 2012).
Integration and Nonintegration of Regimes
Governments have regulated oligopolistic industries with integrated regimes that establish rules for market participants at the global level. However, more unintegrated regimes are developed to govern competitive industries. Governments have a low rate of making agreements outside the treaty regime established to govern an industrial sector, but they have a much higher rate of making these unintegrated agreements to govern competitive industries. In regulating an oligopolistic industry, governments tend to economize on transaction costs by making changes to an existing regime. Although governments also make changes to regimes dedicated to governing competitive markets, they tend to make new agreements or to utilize existing independent institutions or agreements to govern the issue.
Figure 2 shows the rates at which governments have made integrated and unintegrated agreements in issue-areas stemming from industry, from agriculture or land use, from both sectors, or from neither. These aggregated categories are meant to provide a broad-ranging, macroscale estimate of how sectors with different market structures have been governed globally. The nineteen treaty processes were coded in terms of the ratio of industry-related to land-related participants.7 This measure includes treaty processes that govern both industrial and agricultural forms of production (e.g., climate change agreements) in the Both Sectors category.
Treaty processes that govern industrial sectors (e.g., refrigerants, chemicals, hazardous waste, or shipping) have nearly four times the rate of integrated agreements as treaty processes that govern agriculture or land use sectors (e.g., forestry, seeds, or land cultivation). The former processes also have half the rate of unintegrated agreements of the treaty processes that govern land-related issues. If we consider treaty processes that govern both the industrial and land use sectors, the gap in the rates of unintegrated agreements between processes regulating industrial sectors and processes regulating both sectors is even wider, reflecting that issue-areas such as climate change or international rivers tend to create significant challenges for governments, which prefer to economize on the transaction costs of global governance.
A regression analysis substantiates these results when other relevant variables are considered, including the economic and population growth of developing countries, the decision-making rules of the regime, the number of parties to the regime, and the number of sovereign states in the world. In particular, results from stratified Cox proportional hazards models indicate that governments have made integrated agreements more frequently in industrial than in agricultural and land sectors, accounting for the effects of the other variables mentioned above.8 Conversely, governments have made unintegrated agreements with a higher frequency in the context of agricultural and land use sectors than of industrial sectors.
Legally Binding and Nonbinding Agreements
There is a similar divergence in the legalization of governance. Specifically, governments have been more likely to legalize rules for market participants in industrial than in agricultural sectors. Nonbinding agreements are more frequently made for issue-areas that pertain to agriculture or land use than for industrial sectors. There is a high correlation (65 percent) between the frequencies of legally binding agreements and integrated agreements, which is complemented by the 40-percent correlation between the frequencies of nonbinding agreements and unintegrated ones (p values < 0.05).
More generally, the relationship is strong between the dominance of trade groups at meetings and the prospects of a legal agreement. Specifically, the treaty processes with the most frequent nonbinding agreements tend to have the lowest concentrations of trade groups at their meetings. However, the treaty processes with the highest concentrations of trade groups tend to rarely have nonbinding agreements. When we consider legal agreements, the relationship is reversed: The regimes with the highest concentrations of trade groups have the most frequent legal agreements but the least frequent nonbinding agreements. These relationships are evident in Figure 3, where the difference between the low and high ends of the trade ratio scale is statistically significant, in the case of nonbinding agreements (p value < 0.05).
Commitments Made Globally or Nationally
Finally, governments have been more inclined to make global standards and deadlines in regulating oligopolistic industries than in regulating competitive ones, much as they have been more inclined to make integrated and legalized agreements to govern the oligopolistic industries than to govern the competitive ones. The energy, agriculture, and forestry sectors have usually been handled on a case-by-case basis at national or subnational scales, despite the fact that a global treaty was in place with a direct or an indirect mandate. This reflects the difficulty for governments of eliciting technical and practical adjustments by businesses in competitive industries using international agreements, as compared to their preference for standardized global rules applying across different national markets when regulating oligopolistic industries.
Historically, governments have regulated oligopolistic industries with global rules or global standards at a higher rate than they have regulated competitive industries with such rules or standards. Shipping, industrial chemicals, and even agricultural chemicals (which are manufactured by oligopolistic corporations) have all been regulated with rules and standards that apply to all countries, even if the rules provide for a lagged compliance period for certain groups of countries. The commitments were not left up to governments to develop independently at the national or subnational level.
By contrast, governments have often allowed member states to regulate competitive industries with bottom-up, nationally specific commitments. The main exception has been the regimes handling plant genetic resources, and in particular liability and transnational exchange issues, which nonetheless concern multinational corporations operating as oligopolies. Issues such as logging, farming practices, and land-related biodiversity concerns have been left up to governments to decide, within the broad rules or guidelines specified in the global regime, effectively making those commitments nationally specific and nationally determined.
Figure 4 indicates that the frequency of agreements providing for national commitments (or made on a subglobal scale) reflects both the ratio of trade groups to all groups and the sector(s) under regulation. In particular, subglobal agreements are less frequent when governing issues stemming from industrial sectors, where trade groups tend to dominate. By contrast, subglobal agreements are far more frequent when governing issues stemming from agricultural and forestry sectors, where trade groups are much less prevalent at the global level. Cox regression analysis substantiates these descriptive patterns.
Despite their political influence, oligopolies have been more stringently regulated than competitive industries in global environmental governance. Although oligopolies may have more influence in blocking or shaping regulations, the same market structures and capital resources that enable them to wield political influence also enable them to transform markets through technological innovation and practical adjustments, without the need for intensive and extensive government intervention in the market. Consequently, oligopolies are targeted with relatively stringent regulations, but competitive industries are not expected to meet similar requirements. Global regimes are designed to fit the market, because market structures constrain the implementation and cost-effectiveness of global regulations.
In protecting the ozone layer, governments have regulated chemical industries differently, depending on their intermediate and downstream markets. Businesses in the industrial sectors have been expected to meet consistent phase-out deadlines, without much expectation that they would need additional time to complete the necessary technological conversions to meet legal commitments. The agricultural application of ODS chemicals has been the subject of controversy, however, as exemption requests by wealthy countries became the norm in certain farming applications.
Similarly, in designing the new mercury regime, governments adopted a different regulatory approach for artisanal gold mining than for industrial sectors. Legally binding deadlines were defined for industrial sectors, and these deadlines applied to the entire sector in a member state. The phase-out requirements for mercury-added products or processes did not vary by country, but countries are expected to vary in reducing mercury use in artisanal gold mining. The regime is designed to provide a single set of legally binding targets for industrial sectors, and precisely the opposite expectations for artisanal gold mining.
A quantitative analysis indicated that regimes with the highest concentrations of trade associations at global environmental meetings are more often governed by legally binding, integrated, and globalized rules. By contrast, the treaty regimes with the lowest concentrations of trade associations at the meetings are more often governed by nonbinding and unintegrated agreements that provide for country- or region-specific governance. Consistent with the regulation of gold mining and industrial sectors under the mercury regime and the regulation of industrial and agricultural chemicals under the ozone regime, the quantitative analysis also indicated that the treaty regimes governing industrial sectors experience more integrated, globalized, and legalized models of regulation than those governing agricultural and land use sectors.
These findings add to the environmental politics literature by demonstrating that market structures affect regulatory possibilities, not only business power. Future research should elaborate the policy instruments that emerge for governing competitive land use industries and the conditions for exporting models by which to regulate those industries across geographic settings. This would add to the literature on the political economy of global governance by further defining the constraints on regime design and the conditions for improved governance.
For example, the Herfindahl-Hirschman Index is used in United Sates competition law, with particular values denoting how competitive (or noncompetitive) an industry is, but that index is based on the market shares of producers. Index values above 2,500 indicate that an industry is highly concentrated and likely noncompetitive.
Author’s interview with DuPont representative, Bali, Indonesia, 25 July 25, 2012.
Author’s interview with the Artisanal Gold Council, phone interview, April 20, 2015.
Author’s interview with the World Chlorine Council, phone interview, February 8, 2013.
Author’s interview with UN mining consultant, phone interview, April 11, 2015.
The Law of the Sea, the International Watercourses Convention, the London Dumping Convention, and the FAO treaty on illegal fisheries do not have LOPs. Nonetheless, the International Maritime Organization administers the London Dumping Convention, and the LOPs from the Marine Environment Protection Committee of the International Maritime Organization can be used as a proxy.
The online appendix provides coding information for these sectoral categories.
The online appendix provides modeling details and results.
I thank Robert Keohane, Christina Davis, and Helen Milner for feedback on this project. I also thank audiences at Duke University, the University of California at Santa Barbara, the University of Massachusetts Boston, Columbia University, the University of Colorado Boulder, and the University of Georgia, and commentators at the 2013, 2014, and 2015 Annual Conventions of the International Studies Association. I am especially grateful to three anonymous reviewers and the editors of Global Environmental Politics for insightful feedback.