Treaty Ratification

Countries ratify treaties to participate in international institutions. To begin with, structural factors may influence a country's basic propensity to ratify treaties. Economic openness may increase a government's incentive to ratify treaties for reputational reasons,Footnote ¹² while economic wealth may make countries willing to pay for the ‘luxury good’ of environmental co-operation.Footnote ¹³ Similarly, previous research highlights the importance of domestic political institutions. Institutional constraints, such as numerous veto players, may impede ratification,Footnote ¹⁴ and democracies have greater incentives to provide public goods to voters participating in international institutions.Footnote ¹⁵ In autocratic societies, treaty ratification may also be a small concession to domestic opposition.Footnote ¹⁶

Scholars have also recognized the importance of domestic constituency preferences. Nuanced studies of the strategy of treaty ratification emphasize the benefits that accrue to the government and influential domestic constituencies. For example, Simmons and Danner argue that while some governments ratify the International Criminal Court for ‘expressive’ reasons, others do so as a credible commitment to violence reduction.Footnote ¹⁷ Similarly, democratizing states sought ratification of the European Convention for the Protection of Human Rights and Fundamental Freedoms in the aftermath of World War II as a credible commitment to human rights and democratization.Footnote ¹⁸

Other studies examine the influence of organized interests. For example, trade agreements are largely shaped by their expected effect on companies’ profits.Footnote ¹⁹ In environmental politics, Bernhagen argues that influential business groups generally undermine the ratification of environmental treaties because they prefer to avoid new regulations.Footnote ²⁰ Conversely, Spring and Vaahtoranta emphasize the importance of the net environmental benefits that a state would obtain from treaty co-operation.Footnote ²¹ Simmons has also argued that domestic activists often mobilize in support of ratifying human rights treaties, and thus induce the government to pay the sovereignty cost of participation.Footnote ²²

International factors, such as competition and diffusion, are also important. Elkins et al. write that the ratification of bilateral investment treaties is driven by competition for capital,Footnote ²³ while Von Stein emphasizes the importance of flexibility provisions for participation in climate treaties.Footnote ²⁴ Another important factor is peer pressure. Neumayer finds, for instance, that regional peer pressure has influenced the ratification of the Second Optional Protocol to the International Covenant on Civil and Political Rights.Footnote ²⁵ With regard to power politics, McLean and Stone argue that the European Union has exercised regional hegemony to induce Eastern European countries to ratify the Kyoto Protocol.Footnote ²⁶ Drezner notes that regulatory convergence at the international level is most likely when major powers agree on what kind of regulation is in their best interest.Footnote ²⁷

Regulatory Co-operation

Businesses usually oppose new regulations aimed at reducing negative externalities, such as pollution.Footnote ²⁸ Since regulations require companies to make costly changes to their business practices, and such changes are costly,Footnote ²⁹ most companies prefer to avoid them. This notion is widely accepted in the literature on global environmental governance, and many scholars have emphasized the importance of business opposition to environmental regulations in explaining policy outcomes.Footnote ³⁰ Though most such studies are silent on treaty ratification, Bernhagen argues that business groups usually oppose environmental treaty ratification because such treaties generally ‘require firms to reduce emissions, increase recycling, pay more for energy, or switch to alternative, often more expensive, fuels and inputs’.Footnote ³¹

However, businesses may sometimes ally with activists who support new regulations because regulation creates rents for incumbents through barriers to entry.Footnote ³² In the international context, Vogel has argued that import competitors may benefit from regulations that implicitly discriminate against foreign producers.Footnote ³³ According to Kono, democratic governments may use such standards to ‘obfuscate’ protectionist policies from the general public.Footnote ³⁴ DeSombre also maintains that even if businesses initially oppose domestic regulations, they may ultimately accept them as a fait accompli and demand that the government induces other countries to adopt similar regulations.Footnote ³⁵

While this literature has largely ignored technology innovation, there are some notable exceptions. BenedickFootnote ³⁶ and ParsonFootnote ³⁷ argue that a key reason why the United States supported the Montreal Protocol on ozone depletion was that the American chemical industry expected to be able to use inexpensive substitutes for ozone-depleting chemicals in production.Footnote ³⁸ Clapp argues that industries may support the negotiation of environmental treaties when they do not compromise valuable intellectual property rights and create markets for new products.Footnote ³⁹ However these studies do not take into account the effect of technology innovation on national ratification, or conduct empirical hypothesis tests.

Global Regulation, Technology Innovation and Firm Profits

The effect of global regulation on profits depends on a firm's technological innovativeness. Global regulations restrict the use of existing products, with two effects. First, producers of inexpensive, simple products may lose their market share. Secondly, new opportunities emerge for marketing advanced products that comply with the global regulations.

The more that a firm's core competence lies in the production of inexpensive goods, the less able that firm is to exploit the new markets created by regulations.Footnote ⁴¹ Lacking the capacity to produce new technology, firms that mostly compete on price lose their competitive advantage as the simple products they were previously able to produce at low cost are banned.

Conversely, the more active a firm is in research, and the more it seeks to protect valuable new intellectual property through patents, the more it can profit from the new markets that regulations create.Footnote ⁴² As regulations prevent the use of existing products, whether for environmental or health concerns, the consumers of these products seek alternatives. Such alternatives must meet the new regulations, so product innovation is necessary. Firms with considerable innovative capacities invest in research and development (R&D). In doing so, they increase their market share because their competitors, which are not as innovative, cannot offer inexpensive alternatives to these new products, provided that intellectual property rights are adequately protected.Footnote ⁴³

Incentives to innovate within the regulated industry are another key factor. As Levin et al. show in a major survey of multiple manufacturing industries, the importance of technology innovation, and the ‘appropriability’ of profits from it, varies substantially across industries.Footnote ⁴⁴ Companies in relatively mature industries with slow technology innovation may not be in a position to develop highly profitable products in response to new environmental regulations, which may condition possibilities for corporate support of international regulation.

The Role of Treaty Ratification

If national industries expect profits from global regulation, what can the government do? A government's decision to ratify can help a treaty enter into force. Treaty ratification also puts normative and reputational pressure on other states, especially ‘peers’ with similar characteristics.Footnote ⁴⁵ Ratification increases the cost of non-compliant behavior through reputational and other enforcement mechanisms.Footnote ⁴⁶ By ratifying a regulatory treaty, officials can also participate in decision making on implementation, enforcement and monitoring. Additional ratifications strengthen regulatory treaties, so technologically innovative industries have incentives to support their ratification.

Governments are interested in a national industry's profits when this industry is economically important and politically influential. If the companies in the industry have similar interests, they can engage in collective action to shape policy; even if their interests diverge, a dominant company may succeed in lobbying. While many of the most advanced technology industries are now thoroughly transnational, many political channels of influence remain national – or regional in the case of the European Union.Footnote ⁴⁷ Thus national legislatures have a key role to play in the politics of treaty ratification.

For multinational companies, the picture is also complicated by foreign subsidiaries. Subsidiaries with little innovation capacity may lobby on behalf of their innovative parents, so producers in one country may lobby for new regulations even if their own innovation capacities are limited. Even then, their interests are often aligned because the production of technologically advanced products calls for a highly skilled workforce, regardless of whether most of the R&D is conducted abroad. Thus subsidiaries have similar incentives as their parents unless the former do not participate in the production of advanced technology.Footnote ⁴⁸

Industries can lobby for treaty ratification in several ways. Public communication allows industry associations to argue their case for or against treaties. For example, major biotechnology companies and trade associations initiated a public campaign to support the ratification of the Rotterdam and Stockholm Conventions. On 13 July 2004, Mr. Steven Goldberg of CropLife America, a biotechnology industry association, testified to the House Committee on Energy and Commerce, noting that the organization's ‘member companies develop, produce, sell and distribute virtually all the crop protection and biotechnology products used by American farmers’ and arguing that ‘the expedient US ratification and implementation of these treaties is vital to protect our country's interests’.Footnote ⁴⁹

Policy makers respond to lobbying for several reasons. In addition to providing direct political support, such as campaign funding,Footnote ⁵⁰ industries can play a useful role due to their informational advantage.Footnote ⁵¹ They hold private information regarding the economic consequences of regulatory treaties, and they do not have incentives to support a treaty that would reduce their profits. While the industry has its own incentives to misrepresent information, at the very least the government can trust that the industry offers political support for treaty ratification. If an innovative industry supports treaty ratification, it also has incentives to participate in the ratification and implementation process.

Empirical Implications

We expect that nations with substantial technology innovation capabilities have strong incentives to ratify multilateral environmental treaties.

In addition to this primary hypothesis, our argument has empirical implications for institutional design. If regulatory treaties work when industries support them, it seems reasonable to expect that they are also designed to garner industry support. In particular, such treaties should create new markets for safe and sustainable products while otherwise imposing a limited regulatory burden on businesses.

Technology Innovation in the Pesticide Industry

While the production of agrochemicals, including pesticides, is a genuinely global industry, the largest firms are clustered in a few countries. Table 1 lists the ten largest agrochemical firms, per 2007 sales revenue.Footnote ⁵⁴ The U.S.-based chemical giants Dow and DuPont are the only other agrochemical firms to rank among the world leaders in R&D activity, though both rank lower than their German counterparts with respect to agrochemical sales and overall R&D activity.

Table 1

The Ten Largest Agrochemical Firms in 2007

Note: numbers in parentheses denote the firm's rank on the IEEE's 2007 ranking of the world's 100 largest R&D firms (if applicable).⁵⁵

While detailed data on national industry characteristics are not available for many countries, it is nonetheless useful to examine the composition of national biotechnology industries in broad strokes. The Organisation for Economic Co-operation and Development (OECD) has collected recent data on biotechnology industries in twenty-six countries, as shown in Table 2. In addition to biotechnology patenting rates, the table shows (1) the number of firms with a primary focus on biotechnology per million inhabitants, (2) the biotechnology R&D intensity of the country and (3) ratification dates for the Rotterdam and Stockholm Conventions. If our argument is correct, then increases in the number of companies and R&D intensity should be negatively correlated with ratification delay.

Table 2

Biotechnology Industry, Technology Innovation and Treaty Ratification in Select Countries

The data in Table 2 show that biotechnology patents are strongly correlated with the number of biotechnology firms and R&D intensity. First, the correlation between the number of patents and the number of firms is r = 0.51, while the correlation between the number of patents and R&D intensity is as high as r = 0.81. Secondly, the table shows a clear pattern: even within the set of mostly advanced industrialized countries, there is a strong positive association between biotechnology innovation and treaty ratification. With one notable exception (the United States), treaty ratification occurs faster in countries that have active biotechnology industries. As expected, the correlations between biotechnology patents and the year of ratification for the Rotterdam and Stockholm Conventions are negative: r = −0.22 and r = −0.23, respectively.

The Rotterdam Convention

The Rotterdam Convention was adopted on 10 September 1998 and entered into force on 24 February 2004. It requires exporting parties to obtain explicit consent from importing states before shipping hazardous substances (enumerated in Annex III). Parties must also notify the treaty secretariat whenever they impose significant domestic regulations on chemicals not already governed by the treaty (Article 5). If a new pesticide or chemical is regulated by parties from two different regions, that substance is forwarded to a review committee for possible inclusion in the treaty (Article 5). New substances may also be proposed for inclusion in the treaty at the request of a single developing country or economy in transition, if the party has experienced significant problems from the use of that substance (Article 6). However the ultimate decision about whether to include a new substance in treaty regulations is left to the Meeting of Parties to the Convention, which operates by consensus. While it is relatively easy to initiate the consideration of additional regulated substances, the treaty creates a number of veto points that may block such additions.

The requirement of consensus to add new chemicals is consistent with our theory of corporate preferences in international regulation. Industry may favor regulation if it creates markets for more profitable new products. However, industry will then have a strong incentive to protect those new products from potentially costly regulation. The design of the Rotterdam Convention reflects these incentives well. While it does not ban the production of older, ‘generic’ pesticides and chemicals, the PIC mechanism substantially increases transaction costs for firms that wish to sell such products internationally. The Rotterdam Convention also represents an implicit judgment that some pesticides and industrial chemicals are superior (at least environmentally) to their competitors. Thus the treaty may confer valuable reputational benefits for firms.

At the early stages, industry groups opposed a strong treaty. The Chemical Manufacturers Association (CMA), backed by a coalition including the United States, opposed an EC proposal to modify text referring to ‘acutely hazardous’ chemicals to simply ‘chemicals with severe environmental or health effects’.Footnote ⁵⁶ Later in the meeting, the CMA unsuccessfully backed a proposal to keep legal liability on the side of importing countries, rather than with exporters. However as some of the uncertainties surrounding the form of regulation were resolved, industry opposition weakened. By June 1998, at the final preparatory conference prior to the Rotterdam meeting, the chemical industry was squarely behind the adoption of a legally binding agreement involving prior informed consent. Industry's main interest at this point, as expressed by the representative from the International Council of Chemical Associations (ICCA), was to restrict regulation to the twelve chemicals explicitly designated in the treaty text. The chemical industry favored relatively stronger requirements for the expansion of regulation to new chemicals, arguing in favor of ‘risk assessment’, rather than ‘precautionary’ criteria.Footnote ⁵⁷ Per our theory, industry favored regulations that create new markets but opposed rules that could limit those markets in the future.

The provisions and negotiations undermine the claim that the Rotterdam Convention is an instance of shallow co-operation. Given that it allows the inclusion of new chemicals, ratification would not be an entirely ‘safe bet’ for a country that can easily comply with the initial provisions yet is not interested in deeper co-operation. Similarly, strong industry support for the treaty during the final negotiations is difficult to reconcile with the idea that the Convention is just a scrap of paper: if shallow co-operation is inconsequential, the chemical industry should be indifferent, not enthusiastic, about the treaty.

The Stockholm Convention

The Stockholm Convention was adopted on 22 May 2001 and entered into force on 17 May 2004. It involves commitments to eliminate or restrict the very production or use of persistent organic pollutants (POPs) (Article 3). Specifically, the Convention establishes three types of POPs, with varying degrees of regulation. Annex A to the Convention lists pesticides and industrial chemicals, the production and use of which are to be eliminated. Annex B lists POPs that shall be subject to greater restrictions. When the Convention text was adopted, this category consisted solely of DDT. Parties rejected an outright ban primarily because DDT has valuable applications in combating vector-borne diseases, such as malaria. Finally, Annex C lists pollutants that are produced as by-products of other industrial processes. Although the Convention provides for the eventual elimination of these substances, short-term restrictions are subject to feasibility concerns. The Convention also clearly establishes a procedure for including new pollutants in the future. The Stockholm Convention also includes a number of other notable provisions: measures designed to help developing countries enhance their regulatory capacity, including information exchange (Article 9), technological transfers (Article 12) and financial transfers (Articles 13 and 14).

The negotiation of the Stockholm Convention was less dramatic than that of the Rotterdam Convention. The most contentious issue throughout the five negotiating sessions was that of financial transfers to developing countries. The G-77 and China pressed for the creation of a separate multilateral fund dedicated specifically to building capacity for management of POPs. The majority of industrialized countries preferred to use an existing institution for this purpose – the Global Environment Facility (GEF) – so they could wield relatively greater influence. In the end, negotiating parties compromised by agreeing to use the GEF on an interim basis.

There was little disagreement concerning the initial list of pesticides and chemicals – the ‘dirty dozen’ – that would be subject to the Convention. Indeed, most of these had already been phased out by contracting parties. The ICCA, which participated in negotiations leading up to the adoption of the Stockholm Convention, emphasized its support for the Convention through negotiations. As with Rotterdam, industry's main concern centered on the procedure for including new substances in the Convention. The ICCA consistently argued that new substances should be evaluated by risk assessment, rather than a strict interpretation of the precautionary principle. At the first meeting of the Stockholm Convention negotiating committee, held from 29 June to 3 July 1998, the ICCA argued that ‘the scope of the convention should not be expanded beyond the [original twelve] POPs until science-based criteria’, since risk assessment had been established within the Convention.Footnote ⁵⁸ Subsequently, the ICCA sought to nudge the bar higher for including new chemicals under the Convention. It reiterated its earlier position that the Convention should prioritize the substances that had already been identified, and suggested that ‘in-depth assessments’ must reveal ‘information that clearly indicates that the substance is of global concern’.Footnote ⁵⁹

The final text of the Stockholm Convention reflects this preference. Article 8 establishes the Conference of Parties as the final arbiter for inclusion of new substances. Although Article 8 directs the Conference of Parties to act ‘in a precautionary manner’, it also specifically provides for risk management. Notwithstanding this compromise, the Conference of Parties uses a consensual decision rule. Consistent with our expectations, industry supports the regulation of unprofitable substances in order to create new markets for more profitable substitutes.

Neither the above provisions nor the negotiations are compatible with the shallow co-operation hypothesis. The Stockholm Convention allows the inclusion of new chemicals, and the strong industry support indicates that they perceived the treaty as highly profitable. If industry expected the treaty to be inconsequential window dressing, it would not have expended considerable effort to promote it.

The subsequent behavior of the agrochemical industry shows how technological innovativeness created support for global regulatory treaties. In 2007, only three years after the Stockholm Convention entered into force, BASF launched a major new commercial product in the long-lasting insecticide nets market sector. This product, ‘Interceptor’, is designed to replace the use of DDT as a means of controlling vector-borne diseases, the only application that remains permissible under the Stockholm Convention. Since developed countries have banned the domestic use of DDT, production of the chemical has shifted to developing countries. The eventual inclusion of DDT as a controlled substance under the Stockholm Convention represents one of the major remaining business opportunities for highly innovative manufacturers, and BASF's behavior indicates that this opportunity has not gone unnoticed.

The April 2011 addition of a highly toxic pesticide, endosulfan, to the Stockholm Convention provides further evidence of this dynamic. Endosulfan is not a new product. The German agrochemical firm Farbwerke Hoechst AG began selling it in the United States in the 1950s.Footnote ⁶⁰ By the 2000s, however, concern over adverse health effects associated with endolsulfan led a number of OECD countries to consider bans on the production and use of the substance. Accordingly, major European producers such as Bayer CropScience began to shift their commerical portfolios away from endosulfan.Footnote ⁶¹ By the time endosulfan was approved for inclusion under the Stockholm Convention, over the initial opposition of India, global production had shifted from developed to developing country firms specializing in the production of non-patent-protected chemicals. The reason for India's reluctance to support the ban was not hard to discern; by 2011, Indian firms accounted for more than 60 per cent of the global production and use of endosulfan.Footnote ⁶²

Independent Variable

Our primary independent variable is Biotechnology Innovation Capacity, lagged by one year to avoid possible simultaneity bias. We measure biotechnology innovation capacity by the yearly number of biotechnology patent applications filed under the 1971 Patent Co-operation Treaty (PCT), by country of inventor. A PCT application is interpreted as an application for a ‘priority status’ that prevents other inventors from patenting a given invention in foreign markets. As such, it signals the inventor's intent to commercialize an invention in international markets. The data are from the OECD Patent Statistics database.Footnote ⁷⁰ A major advantage of this dataset is that there are no missing data because the patent applications are centrally administered.

Biotechnology patents are particularly suitable for measuring commercial innovation. The biotechnology category is sufficiently broad to include not only direct biochemical substitutes for older pesticides and chemicals but also innovations in genetic engineering that may reduce the need for pesticides. It is defined by the OECD, so our own subjective biases cannot contaminate the measure. The patent category contains selected patents from five major classes: agriculture, health, organic chemistry, biochemistry and laboratory instruments. From each class, OECD researchers have selected more detailed sub-classes that specifically relate to biotechnology. While it is impossible to verify the exact number of patents that are somehow related to pesticides and other chemicals regulated by the Rotterdam and Stockholm Conventions, we investigated the description of each sub-class and found that all of them contain products that could be used in plant disease control. Additionally, only one of the five major classes specifically focuses on health, the other major use of biotechnology in modern societies, and each sub-class within the health category contains products that could also be used in plant control. As a robustness test, we also included all PCT patents for each country-year in the regression, and found that they do not predict ratification of either treaty, while biotechnology patents continue to exert a positive influence.Footnote ⁷¹

We normalized the number of patents by gross domestic product (GDP) to account for variation in the size of the economy.Footnote ⁷² Specifically, we divided the annual number of patents by GDP, measured in hundreds of billions of constant 2000 USD.Footnote ⁷³ Without such normalization, large countries such as the United States and Germany would completely dominate over all other countries.Footnote ⁷⁴ For example, Denmark is a small country with a population of 5.5 million in 2010, but it has a very innovative biotechnology industry. Without accounting for the size of the economy, the critical importance of biotechnology for the Danish economy would not be reflected in the data. For political lobbying, it is also the relative (not absolute) size of an industry within a country that matters.Footnote ⁷⁵ As a robustness check, we also verified that all of our results hold without logarithmization.

The use of patents as a proxy for innovation capacity is established in the scholarly literature.Footnote ⁷⁶ It yields a quantifiable intermediate measure of innovation output, and given our focus on commercial innovation, the fact that some technologies may not be patentable is a relatively minor concern. Indeed, studies of patenting practices highlight that the chemicals industry is unusually dependent on patents for intellectual property rights protection.Footnote ⁷⁷ To be sure, using patents as a measure of innovation performance is not without disadvantages. As noted in the OECD Patent Statistics database, ‘the propensity to patent differs across countries’, and some patents ‘have no industrial application’.Footnote ⁷⁸ The former problem is mitigated by our usage of international patent applications; the latter concern is minor because we focus on biotechnology innovations that require substantial R&D investments, and are thus rarely forthcoming unless they promise substantial commercial profits.

One issue concerning the use of international patents is the delay between the initial research investment and the patent application. A biotechnology patent filed in 2010, for example, often results from years of research and development. For our purposes, this delay does not present a problem. In all countries with high patenting rates, these rates either remained stable (United States) or increased somewhat over time (Denmark, Switzerland). If anything, the delay thus leads us to underestimate the importance of biotechnology innovation in key countries.

Perhaps the most important impediment to empirical inference is the possibility that biotechnology patents are correlated with other factors, such as GDP per capita or regime type. In addition to controlling for such factors, we conducted several additional tests to reject the spurious correlation hypothesis. First, we examined whether ratifiers and non-ratifiers are systematically different along these lines. As shown in Table 4, the only systematic difference that we found pertained to regime type. Yet when we exclude non-democracies from our dataset, our results remain unchanged. Secondly, we examined correlations between biotechnology patent rates and these variables. The only relatively strong correlation we found was between GDP per capita and biotechnology patent rates, r = 0.52. We thus estimated all our models excluding countries with a GDP per capita less than USD 5,000, and all our results remained unchanged. Finally, we excluded all countries from our dataset that did not produce any patents in any year. Again, our results remained unchanged.

Table 4

GDP Per Capita, Population and Regime Type for Ratifiers Versus Non-ratifiers (by 2011)

Note: the values are for the first possible year of ratification.

Control Variables

We control for a number of additional variables that might influence ratification. The appendix introduces several additional control variables that we added to our main models to scrutinize the robustness of our empirical results. All control variables are lagged by one year to avoid possible simultaneity bias. Those that are not normally distributed are also logarithmized to prevent outliers exerting undue influence on our results.

Because domestic political factors may influence the speed of ratification, we control for the influence of domestic political institutions on treaty ratification. Since treaty ratification often requires legislative approval, we distinguish parliamentary systems from presidential ones, in which the legislature approving the treaty may be controlled by a different party than the executive, which controls the signature of a treaty. Given that the executive and legislative bodies are separated, presidential systems are likely to be slower to ratify treaties than parliamentary systems. Therefore we include in our models a dichotomous measure that distinguishes whether or not a country has a parliamentary system. We supplement this by controlling directly for fractionalization – a measure of divisions within the government, and thus the probability of ratification failure due to conflict – as measured in the Database of Political Institutions.Footnote ⁷⁹ As a final robustness check, we apply a measure of ‘ratification difficulty’ from Simmons.Footnote ⁸⁰ This variable captures institutional impediments to speedy ratification, and thus accounts for supply-side constraints on treaty participation.

We also control for the relative importance of pesticide production in the national economy. We use the value of pesticide exports because domestic production data are not available. Since both treaties limit the free trade and production of these commodities, countries will ratify more slowly if pesticide exports comprise a relatively large share of the national economy. We obtained yearly pesticide export data from the FAO in USD and normalized it by GDP, measured in billions of dollars.Footnote ⁸¹ We logarithmize these values to limit the influence of outliers, though all results also hold without logarithmization.Footnote ⁸² Since countries report these data voluntarily to the FAO, and selection effects may exist, we supplement export data with a dichotomous measure to distinguish whether or not a country fails to report pesticide exports.Footnote ⁸³

Since this measure does little to capture the domestic demand for controlled substances, we also control for the importance of agriculture to a national economy, measured as the value of agriculture and normalized by GDP, as reported in the World Development Indicators. Countries with large agricultural sectors may ratify these treaties slowly due to the higher compliance costs.

The extant literature shows that economic openness promotes the ratification of environmental treaties.Footnote ⁸⁴ We thus control for trade openness, or the total value of a country's trade divided by GDP.Footnote ⁸⁵ Given influential outliers, we logarithmize this variable.

Since environmental preferences and treaty implementation capacity may depend on national wealth, we add logarithmized GDP per capita.Footnote ⁸⁶ The purpose of this control variable is to account for possible wealth effects, such as increased preference for environmental quality among citizens, as immediate economic hardship is mitigated through enhanced productivity. The data are from the World Development Indicators and measured in constant 2000 USD.

Following the literature on democracy and environmental co-operation,Footnote ⁸⁷ we control for democracy (Polity IV). According to the literature, democratic countries should be more likely to support environmental co-operation because their governments have a stronger interest in providing public goods for the citizenry. In some robustness checks, we also enter a dichotomous variable that distinguishes whether or not the national executive represents a right-wing political party, on the assumption that more conservative governments may be less interested in international environmental co-operation in general.Footnote ⁸⁸ Another robustness check accounts for the presence of an environmental ministry.Footnote ⁸⁹

Since countries may ratify regulatory treaties because they expect benefits from capacity building,Footnote ⁹⁰ we include the bureaucratic quality variable from the International Country Risk Guide to account for this possibility. If countries are interested in capacity building, then bureaucratic quality should, all else equal, reduce the likelihood of ratification in any given year.

We control for the possible influence of membership in relevant international organizations by including dichotomous variables for OECD and EU membership. Since the EU expanded during the period that the Rotterdam and Stockholm Conventions were negotiated and ratified, and the acquis communautaire specifically incorporates environmental criteria, we expand our definition of EU members to include non-members during years in which they were formal applicants for membership. While we unfortunately do not have detailed data on prior pesticide regulations, the OECD variable helps address the shallow co-operation hypothesis. Given that OECD countries have much more stringent pesticide regulations than other countries, the OECD variable reflects the effect of prior regulation on ratification propensity.

Next, ‘green’ countries may have strong incentives to ratify multilateral environmental treaties. To account for this tendency, we add a variable measuring the number of multilateral environmental treaties that each country has ratified prior to 1998, the first year of our dataset.Footnote ⁹¹ Based on previous research,Footnote ⁹² one expects countries to ratify more environmental treaties if their citizens have a strong interest in environmental co-operation. This control variable accounts for this possibility. The data are from Bernauer et al.Footnote ⁹³

To control for domestic environmental interest, we also include the ratio of International Organization for Standardization (ISO) 14001 standards to the country's GDP. ISO 14001 is a global management standard that companies can adopt. Recent research shows that they do so in order to signal their environmental commitment to customers.Footnote ⁹⁴ This variable captures domestic environmental interest in another way and further strengthens the robustness of our empirical approach.

To control for norm diffusion, we include a variable measuring the proportion of other countries in each region that has already ratified. According to previous research on norm diffusion, regional ratification should have ‘neighborhood effects’ that promote ratification.Footnote ⁹⁵ Finally, we control for the effect of population on the speed of treaty ratification.

Four robustness checks warrant a brief discussion. First, we consider an extended model with several regional dummies. Secondly, we include a binary variable that indicates whether a country was an original signatory of the convention in focus. Thirdly, we include US and EU foreign aid to account for possible foreign influence. Finally, we add another measure of the strength of the environmental civil society by counting the number of environmental non-governmental organization (NGO) headquarters located in each country.Footnote ⁹⁶

Sample Selection

For robustness checks, we examined additional sub-samples for each dataset. First, we excluded all autocratic and anocratic country-years, as measured by a Polity IV score of less than +5. This ensures that our results are not driven by different causal processes in democratic and non-democratic countries. Secondly, we conducted outlier tests to investigate whether some countries had undue influence on our statistical estimates. For each treaty, we found several potential outliers, as described in detail in the appendix. When we excluded these outliers from our datasets, our findings remained unchanged. To ensure that our findings also hold among developed countries, as a robustness check we excluded all country-years with per capita GDP less than USD 5,000 (in constant 2000 dollars). Finally, we excluded all countries that had never filed a biotechnology patent application with the PCT. While the statistical significance levels decreased somewhat due to the smaller dataset, the main results continued to hold.

Germany

Germany was among the earliest ratifiers of both the Rotterdam and Stockholm Conventions. The Rotterdam Convention became available for ratification in September 1998 and Germany was the twelfth country to deposit its instrument of ratification in January 2001. The Stockholm Convention became available for ratification in May 2001, and Germany ratified it less than a year later in April 2002 (the fifth country to do so). Germany's early ratification is fully consistent with our theory. Germany has a very large agrochemical industry that is technologically among the most sophisticated in the world. This is reflected in international patents data. According PCT data, Germany had the second-most biotechnology patent applications in the world in 2001 (1,055), behind only the United States. Controlling for GDP, Germany's biotechnology patent application rate was eleventh highest in the world. Given the size of Germany's economy, this is a considerable achievement.Footnote ⁹⁹

We conducted a systematic media search of English- and German-language newspapers, and found very little evidence of ratification controversies or industry opposition to either convention.Footnote ¹⁰⁰ Indeed, our most important finding was that the ratification of neither treaty provoked any public controversy whatsoever. One of the few media articles that specifically addressed the global pesticides regime from September 2002 noted that in Germany, ‘the pesticides industry supports the Rotterdam Convention … members of the [agrochemical] industry association Agrar have already in 1989 voluntarily committed to the PIC principle’.Footnote ¹⁰¹ Consistent with the negotiation history and our theoretical argument, the article also notes that Agrar advocates science-based consideration of future changes in the list of pesticides subjected to PIC.Footnote ¹⁰²

Indeed, the German pesticides industry's own rules prevent it from profiting from pesticides that either convention would proscribe. According to the industry association's website, the obligations of membership ‘include the responsibility to only export such pesticides to foreign countries that are legal in OECD countries and thus meet high environmental and security standards’.Footnote ¹⁰³ Thus it is not surprising that the ratification of the Rotterdam and Stockholm Conventions did not provoke the usual industry backlash in Germany. Indeed, as discussed earlier, Bayer CropScience ceased sales of endosulfan in anticipation of its inclusion in the Stockholm Convention.

India

India is among the least co-operative countries with regard to pesticides. Despite the severe negative health effects of unregulated pesticides use for the rural poor, and chemical accidents such as the December 1984 Bhopal disaster, the Indian government has undermined the Rotterdam and Stockholm Conventions. It only ratified the Rotterdam Convention in May 2005 after it had already entered into force, and the Stockholm Convention in January 2006, also after it had entered into force. Even after ratification, India's national implementation policy has been weak. Industry opposition to both conventions has also been apparent. This is not surprising, given that India has a large but technologically primitive pesticides industry.Footnote ¹⁰⁴ According to our theory, industries in countries like India have incentives to oppose new environmental regulations because they are not in a position to profit from stringent regulatory measures that open new markets for sustainable products. Instead, restrictions on pesticides use would undermine their competitive position vis-à-vis manufacturers from advanced industrialized countries such as Germany.

While Germany's ratification process attracted little media attention, India's behavior within the global pesticides regime has been carefully documented in the Indian and international media. Our media search of English-language newspapers and negotiation documents provided several important insights.Footnote ¹⁰⁵ The Indian delegation appears to have been particularly active in the negotiations on the Stockholm Convention. In the first meeting of the Negotiating Committee, June–July 1998 in Montreal, ‘India … said due consideration should be given to the economic viability of alternatives to POPs for developing countries’ while ‘[t]he Indian Chemical Manufacturers Association (ICMA) noted that DDT is still in use for malaria control in India’.Footnote ¹⁰⁶ In the second meeting, in Nairobi in January 1999, ‘India … stressed the need for differentiated timetables for phasing out POPs’.Footnote ¹⁰⁷ In the remaining meetings, India actively but unsuccessfully led developing countries in demanding sizable financial transfers for the implementation of the Stockholm Convention.

The Indian chemical industry was also very active in delaying ratification of the Stockholm Convention. On 8 June 2003, a co-chairman of ICMA's international treaties expert committee, S. Ganesan, argued that ‘[i]f accepted at this juncture, the globally legally binding treaty would be detrimental to the health of the Indian chemical industry’.Footnote ¹⁰⁸ The Indian chemical industry also tried to coax the government to select ‘conditional ratification’ of the Stockholm Convention that would allow it to opt out of any future decisions regarding chemical regulation. When the Indian government finally ratified the treaty three years later, the ICMA announced that it would ‘boycott’ the national implementation of the treaty.Footnote ¹⁰⁹

While India's negotiation moves and ratification of the Rotterdam Convention did not provoke equally acrimonious public controversies, the Indian chemical industry nonetheless opposed the treaty. Within Indian industry, treaties such as the Rotterdam and Stockholm Conventions are perceived less as mechanisms for environmental co-operation and more as de facto efforts to promote European industry. As the chair of the Indian Chemical Council's Committee on International Treaties explained in a public press release, each time the international community has acted to regulate pesticides or industrial chemicals, ‘developing countries lose one widely used inexpensive generic chemical. Stopping its production, sale and exports [results in] in significant loss of invested capital, revenues (including export revenues) [and] loss of employment to developing countries’. The distributional consequences are far different in European countries, where such such regulation ‘involves no cost (but only gains)’, since the EU ‘nominates a chemical to POPs list only after banning its production within EU and after developing substitute chemical’.Footnote ¹¹⁰

Even after ratification, the Indian delegation has adopted negotiation positions to undermine the expansion of the PIC principle. Perhaps most importantly, it has opposed the inclusion of endosulfan, a powerful pesticide used in cotton farming, in the Rotterdam and Stockholm Conventions. For example, in the fourth meeting of the Conference of the Parties for the Rotterdam Convention, in October 2008 in Rome, India (supported only by Pakistan) exploited the consensus decision rule to prevent the inclusion of endosulfan in Annex III of the Convention.Footnote ¹¹¹ In December 2010, the Pesticides Manufacturers and Formulators Association of India and the Endosulfan Manufacturers and Formulators Association argued that the planned ban of endosulfan, actively promoted by the EU, was intended ‘to destroy 70 per cent of the world exports of the product from India … arguments put forward by an international committee for imposing the ban was not based on scientific evidence, but by the brute display of voting power by the members’.Footnote ¹¹²

United States

The United States is a particularly interesting case for our purposes. In 2001, the number of US biotechnology patent applications filed with the PCT (5,869) was more than five times that of Germany, its closest rival.Footnote ¹¹³ The American delegation was also active in the negotiations of both treaties, and despite the replacement of the Democratic Clinton administration with the Republican Bush administration in 2001, the executive continued to support the ratification of both treaties. Nonetheless, the United States failed to ratify either of them. This case, thus, seems to contradict our argument. However we use process tracing to show that the biotechnology industry supported the ratification of these treaties, while the ratification failure was caused by the anti-regulatory preferences of some Republicans in the Senate, combined with the requirement that sixty-seven of the 100 Senators must vote ‘yea’ for a treaty to be ratified.

While the United States opposed the precautionary principle in the negotiations on both treaties, it was ultimately satisfied with the final texts. Moreover, the executive wing has supported the ratification of both treaties. In April 2002, the Bush administration made a public commitment to ratify the Rotterdam and Stockholm Conventions as soon as possible.Footnote ¹¹⁴ Consistent with our theoretical argument, this decision was strongly supported by the main industry association, the American Chemistry Council (ACC). According to ACC website, ‘The US chemical industry – which participated in the negotiations along with environmental organizations – agrees that the new global rules represented by the treaties are an appropriate step in enhancing health and environmental protections’.Footnote ¹¹⁵ Similarly, the Alliance for the Responsible Use of Chlorine Chemistry – a major coalition of chemicals manufacturers and chlorine users – argued in September 2003 for ratification, noting that ‘ARCC support of the Stockholm Convention and related treaties is based on achieving three goals: consideration of the legitimate interests of America's workers and companies; identification and adoption of global environment protections; and the promotion of economic growth for both developed and developing nations’.Footnote ¹¹⁶

Why, then, did the United States fail to ratify the two treaties under both the Bush and Obama administrations? The reason seems to be opposition in the Congress – the Senate in particular – to allowing the Environmental Protection Agency to incorporate international rules into decision making on chemicals. According to Kristin S. Schafer, ‘[t]he primary barrier to ratification has been a reluctance to establish a reasonable domestic system for taking action when new chemicals are added under the treaty’.Footnote ¹¹⁷ While even key Republicans supported the principles underlying the conventions, they were concerned about conflicts between domestic and international law.Footnote ¹¹⁸ Thus attempts to formulate domestic implementing legislation that would allow ratification have continued to fail, despite unusually strong support from environmental, consumer and industry groups. The ratification failure of the United States can be ascribed to domestic political institutions, not industry opposition.