Exogenous Technology

As a baseline, consider a simple game whereby states i = A,B simultaneously decide on co-operation, $$${{c}_i} \in \left\{ {0,1} \right\} $$$ . If at least one state rejects co-operation, so that c_A = 0 or c_B = 0, both states obtain a status quo payoff. Without loss of generality, we normalize it to zero. If both states co-operate, c_A = c_B = 1, they negotiate and ratify a treaty. Each state i acquires a co-operation benefit V_i > 0. This benefit can be thought of as flowing from the eradication of a negative externality, such as transboundary air pollution. However, each state i must also pay a price P_i > 0. This price represents the cost to consumers of co-operation. For example, reducing transboundary air pollution implies economic adjustments that are not costless. We allow both co-operation benefits V_i and costs P_i to vary across states.

Finally, international co-operation creates a new technology market. The profits from this technology market to state i are denoted by α_iπ_i , where α_i > 0 is a coefficient that measures state i's sensitivity to profits. We allow negative profits, π_i < 0, to capture the possibility that state i loses profits because its domestic producers were competitive without co-operation. For example, a state that is heavily dependent on coal production cannot expect to benefit economically from climate co-operation.

If governments are generally interested in technology markets, how can α_i be low? States sometimes do not have any industrial capacity in the relevant issue area. For example, France has a sizable nuclear industry, but has very few companies in the wind power sector.Footnote ⁴² If co-operation on renewables would mostly influence the relative profitability of wind power vis-à-vis solar, without affecting the demand for nuclear energy, the French government would have a low α_i .

More generally, what determines a state's profit coefficient α_i ? It is positively associated with the economic importance and political clout of the industry in focus,Footnote ⁴³ the degree to which this industry produces positive spillovers for other national industriesFootnote ⁴⁴ and the influence of national bureaucrats with techno-nationalist preferences.Footnote ⁴⁵

Technology Competition

If both states co-operate, c_A = c_B = 1, the prices P_A , P_B and the profits π_A , π_B are determined endogenously through technology competition. Specifically, each state i selects a technology subsidy s $$$_i} \in [0,\,\infty ) $$$ to domestic firms for the development and commercialization of the new technology. The cost of subsidy s_i to state i is denoted by $$$\frac{1}{2}{{\beta }_i}s_{i}^{2} $$$ , where β_i > 0 is a coefficient that measures the marginal cost of subsidization to state i.Footnote ⁴⁶ For example, it could reflect fiscal constraints, the productivity of the national innovation system,Footnote ⁴⁷ or consumers’ aversion to increased taxes or government debt.

Each state can try to increase its market share through industrial policy, modeled as a subsidy for concreteness. These subsidies reduce the price of the new technology required for co-operation, yet technology competition also has a zero-sum element, in that states compete for markets. Specifically, the subsidy will have the following effects. First, suppose profits to state i depend on subsidies, so that π = π(s_i ,s_j ). While we do not endogenize other determinants of profits here, our model extension below shows that the results continue to hold even if profits depend instead on private investment in the new technology. On the one hand, we assume that π is increasing but strictly concave in s_i : subsidies allow state i to increase market share but with decreasing returns on the margin. On the other hand, we assume that π is decreasing but strictly convex in s_j : if the foreign state j increases subsidies, profits to state i decrease.Footnote ⁴⁸ To simplify exposition, we assume that π(s_i ,s_j ) < 0 when s_i →0 and s_j →∞. If a country completely loses technology competition, it obtains negative profits. Co-operation will inevitably have negative effects on industries producing technologies that will become outdated due to international co-operation. Thus if a state obtains a negligible share of the new technology markets, profits decline.

To avoid implausible outcomes, we let $$$ \pi \left( {{{s}^\ast}, {{s}^\ast} } \right) \gt \pi ( {{{s}^{\ast\ast}}, {{s}^{\ast\ast}} } \right) $$$ if and only if $$${{s}^\ast} \gt {{s}^{\ast\ast}} $$$ . If both states select identical subsidies and increase them proportionally, profits increase. Subsidies generally reduce the production cost of the new technology; as long as market competition is not perfect, in which case all profits would dissipate, this possibility generally allows companies to enhance their profits. Given the importance of economies of scale for new technology commercialization, the assumption of imperfect competition seems plausible.

Secondly, suppose that the consumer price P_i = P_i (s_i + s_j ) depends on subsidies given by both states. Specifically, let the function be decreasing and strictly convex: subsidies reduce consumer prices. Both domestic and foreign subsidies are assumed to have identical effects for analytical tractability. For plausibility, we assume P_i is positive everywhere, so that international co-operation is costly. We also assume P_i (0,0) > V_i and $$${{P}_i}\left( {\infty, \infty } \right)\, \lt\,V_i$$$ , which states that for high (low) enough subsidies, the domestic price of co-operation is above (below) the benefits.

Exogenous Technology

Consider first the case of exogenous technology, so that V_i , P_i , απ_i are all fixed for each state i. Since state i can unilaterally prevent international co-operation, we only need to investigate whether an equilibrium exists that allows co-operation. If it exists, it must be on the Pareto frontier. The necessary and sufficient condition is simply that for i = A,B:

$$ {{U}_i}\, = \,{{V}_i}\,{\rm{ - }}\,{{P}_i}\, + \,{{\alpha }_i}{{\pi }_i}\,\geq \,0. $$

The total benefit, V_i + απ_i , must exceed the consumer cost P_i . States co-operate when both expect benefits from doing so. If this condition fails for at least one state, co-operation is not possible.

Technology Competition

To allow for technology competition, the game must be solved by backward induction. To begin with, suppose both states have agreed to co-operate, $$$c_{A}^{\ast} = c_{B}^{\ast} = 1 $$$ . In this case, both states i simultaneously choose subsidies s_i ,s_j . Each state i selects s_i to maximize its expected payoff, holding s_j constant. The technical details are provided in the mathematical appendix.

The equilibrium subsidies $$$s_{i}^{\ast}, s_{j}^{\ast} $$$ induce an expected technology competition outcome, so the co-operation condition can be written as:

$$ U_{i}^{\ast} \, = \,{{V}_i}\,{\rm{ - }}\,{{P}_i}\left( {s_{i}^{\ast}, s_{j}^{\ast} } \right) + {{\alpha }_i}\pi \left( {s_{i}^{\ast}, s_{j}^{\ast} } \right){\rm{ - }}\frac{1}{2}{{\beta }_i}{{\left( {s_{i}^{\ast} } \right)}^2} \geq 0. $$

States co-operate when they expect net benefits, which depend on the nature of the resulting technology competition.

Nature of Technology Competition

To begin with, let us consider the nature of the technology competition given previous decisions to co-operate, $$$c_{A}^{\ast} \: = \:c_{B}^{\ast} = 1 $$$ . In equilibrium, each state i selects the subsidy $$$s_{i}^{\ast} $$$ to maximize its expected payoff. How does this equilibrium subsidy $$$s_{i}^{\ast} $$$ change with the value of profits (as measured by the coefficient α_i ) and the cost of subsidization (the coefficient β_i )? It is useful to first investigate the best responses of the two states. If the foreign state j increases its technology subsidy s_j , would state i prefer to respond by increasing or decreasing its own subsidy s_i ? Suppose that state i has little interest in market profits, so that α_i obtains a low value. State i's best response would have a positive slope: if foreign state j increases the subsidy, then state i has fewer incentives to subsidize technology. After all, the foreign state is already bringing the consumer price P down by choosing a high subsidy. When state i has a keen interest in market profits, so that α_i obtains a very high value, exactly the opposite is true. When the foreign state j increases its subsidy s_j , state i has an incentive to respond by increasing its subsidy s_i . In this case, the best response has a negative slope. Any increase in the subsidy of one state amplifies technology competition.

Given these observations, we are left with three main scenarios. First, it is possible that both states have increasing best responses, so that technology competition is mutual. Secondly, it is possible that neither state has an increasing best response, so that technology competition does not exist. Finally, one state may have an increasing best response, so that the technology competition is unilateral. These scenarios are illustrated in Figure 1.

Fig. 1

Best responses

Note: on the left, only state A ascribes a high value to profits. In the center, states A and B both ascribe a high value to profits. On the right, both states A and B ascribe a low value to profits.

These observations yield the following comparative statics.

The effects of changing the market value α_i or the technology cost β_i depend only on the foreign state's best response. If the foreign state has a positive best response, increasing α_i or decreasing β_i will induce both states to increase subsidies, generating a ‘race to the top’. If the foreign state has a negative best response, increasing α_i or decreasing β_i will induce state i to increase subsidies, whereas the foreign state j will respond by selecting a lower subsidy.

The case studies below illustrate this notion. In the case of ozone depletion, both the American and British chemical industries were confident that they would be able to develop profitable substitutes for ozone-depleting chlorofluorocarbons (CFCs), and each invested heavily in R&D.Footnote ⁵⁰ In the case of European renewable policy, countries with little initial interest in renewable energy (such as Italy and France) did not react to the EU's Kyoto commitments by subsidizing renewables, while the enthusiasts (notably Germany and Denmark) did so. Through industrial policy, these enthusiasts were able to build internationally competitive export industries that benefited from the expansion in demand caused by the Kyoto Protocol.

What about the payoffs from the technology competition subgame?

If state i becomes increasingly interested in the profits of domestic companies, its payoff in the technology competition subgame increases (decreases) whenever the foreign state is (not) competitive. Since state i suddenly becomes more interested in profits, the value of international co-operation for it depends on whether it is profitable to co-operate internationally. If it is technologically competitive, it will be more keen to engage in international co-operation. Otherwise, international co-operation will face increasing opposition from domestic industries. Next we consider the foreign state's market valuation.

This proposition shows that if the foreign state j's technology competitiveness or incentive to capture market share increases, state i benefits when it is primarily interested in consumer price. In this case, state i has nothing to lose from the resulting increase in foreign subsidies: consumer prices decrease, and the loss of market share is hardly of interest to state i, so the net effect is unambiguously beneficial.

Possibility of Co-operation

In the previous subsection, we characterized the technology competition subgame. Given that technology competition inevitably follows mutual acceptance of co-operation, the nature of technology competition will also affect the probability of co-operation. How do changes in the relevant parameters, such as the subsidy cost β_i , influence the possibility of co-operation? We approach this question by first focusing on technologically symmetric states: if both are equally competitive, what are the conditions for successful co-operation? Next we examine the case of technologically asymmetric states: if one of them dominates the technology competition subgame, is co-operation easier or more difficult?Footnote ⁵¹ First we consider the simple case of symmetric countries.

Under conditions of approximately equal competitiveness, international co-operation is possible so long as both states are primarily interested in consumer prices and technology subsidies are inexpensive but efficacious. Since both states invest heavily in the new technology, the cost of co-operation decreases.

The consequences of techno-nationalism are less obvious. When two symmetric states are interested in national profits, the possibility of co-operation depends on the nature of technology competition. When market profits are equal and positive, international co-operation is mutually profitable. When equal market profits are negative, international co-operation will be too costly for domestic companies, and so it fails. Contravening previous research,Footnote ⁵² the fact that states are techno-nationalist and concerned about relative market share need not impede co-operation.

Pesticide regulation offers a useful illustration of this proposition. The 2001 Stockholm Convention bans a ‘dirty dozen’ of older pesticides that are rarely used in industrialized countries. Consistent with the proposition, the chemical and biotechnology industries in the United States and Europe lobbied for the convention in negotiations, and later pushed their national governments to ratify it. The convention created markets for advanced chemical pesticides and their biotechnological alternatives in developing countries.Footnote ⁵³ Given that both the United States (Monsanto, Dow Chemicals) and Europe (Bayer, Syngenta) had major agrochemical companies, the model correctly predicts industry support from both sides of the Atlantic.Footnote ⁵⁴ We next consider asymmetric states.

Co-operation is possible whenever the technological underdog is not interested in market profits. When it is not interested, it engages in international co-operation because it expects the other state to reduce the consumer price to an acceptable level. When it is very interested, exactly the opposite is true: the prospect of losing in the technology competition induces a state to refrain from co-operation.

This proposition thus captures the idea of juste retour in technology competition. As Sandholtz argues, a key requirement for technology co-operation between states is that they can expect their fair share of the profits.Footnote ⁵⁵ The proposition shows that this is also true when technology competition is a consequence of co-operation on a related problem, such as environmental pollution. The expectation of technology competition may induce countries to reject a treaty because they worry about the distributional effects of co-operation on technology markets.

Tucker offers an interesting illustration from European defense co-operation.Footnote ⁵⁶ As long as the French aeronautics industry dominated its West German competitors, Franco-German co-operation on new fighter aircrafts was possible because neither side was concerned about unfavorable shifts in their international competitiveness. The 1968–1975 Alpha Jet project – led by the French manufacturer Dassault – was an important success, whereas the 1983–1985 European fighter aircraft project failed, because the West Germans demanded a larger share of gains while the French worried that collaboration would allow the rapidly advancing Messerschmitt-Bölkow-Blohm to reach technological parity with Dassault.

These findings are summarized in Table 1, which shows that under symmetry, both states must expect substantial benefits from co-operation, either because (i) each acquires substantial profits from new technology markets or (ii) they are not concerned about technology competition but rather focus on consumer prices. Under asymmetry, however, it suffices that more competitive states are also interested in profits from the new technology markets.

Table 1

The Possibility of Co-operation for Various Parameter Values

Note: under symmetry, both states must expect substantial benefits from (i) eradicating the negative externality or (ii) enhanced profits. Under asymmetry, the less competitive state must be willing to accept low profits.

Model Extensions

In the main model, (i) states select the technology subsidies, (ii) both of them must approve the treaty for co-operation to begin and (iii) neither state can threaten to act unilaterally to solve the collective action problem. What if we relax these three assumptions?

First, suppose private companies are fully responsible for new technologies, so that the state is passive. In this modified model, a private company in each state i would select a technology investment – a private equivalent to a subsidy in the main model – denoted by $$${{s}_i} \in [0,\infty ) $$$ . The private company in state i would maximize some payoff function $$${{F}^i} ({{s}_i},{{s}_j})$$$ , which we leave implicit for generality. In equilibrium, the private companies would obtain payoffs $$${{F}^i} (s_{i}^{\ast}, s_{j}^{\ast} ) $$$ . The expected payoff to state i would be represented by Expression 2 (above). The only difference would be that the states are no longer in direct control of the technology competition outcomes after the treaty enters into force. Thus the model is fully applicable even if states are expected to adopt a ‘hands-off’ approach to technology innovation in the private sector.

Secondly, assume that state A can impose co-operation on state B, for example because it is much more powerful than state B. Alternatively, states A and B could represent country coalitions within a multilateral organization. In this case, voting rules other than unanimity could favor one of the coalitions, perhaps because it has more members. This extension would not compromise the main results, because the competition subgame would remain unchanged. The sole difference would be that at the co-operation stage, only state A would have to select $$$c_{A}^{\ast} = 1 $$$ for the treaty to enter into force. It would suffice to verify that state A may expect to benefit from co-operation.

Thirdly, consider the possibility of unilateral action. Suppose that by paying a fee F > 0, state A could unilaterally enact a policy that is de facto equivalent to a treaty. For example, the United States could unilaterally regulate fisheries imports from foreign countries, and the new regulations could diffuse to foreign countries through trade channels.Footnote ⁵⁷ In this case, the outcome of the game would obviously be unchanged if international co-operation were possible in the first place, as state A would not benefit from unilateral action. But if the fee F were low enough, state A could force state B to engage in collective action if co-operation were failing, because state B alone would expect a negative payoff from the technology competition subgame. Thus the effects of unilateral action would be largely identical to the power asymmetry case examined above.Footnote ⁵⁸

Symmetric Agreement: The Montreal Protocol

Symmetric technological development of CFC substitutes played a major role in the successful conclusion of the Montreal Protocol. States were concerned about the effect of the ozone regime on the market share and considered their technology development level during the bargaining process.

The negotiations, which officially began at a United Nations Environment Programme Governing Council meeting in 1981, resulted in the signing of the Montreal Protocol in September 1987. The agreement called for a 50 per cent reduction in the consumption of certain CFCs by 2000. Although the negotiations were complicated by disputes over scientific uncertainties surrounding the impact of CFCs on ozone depletion, and by the chemical industry's opposition, the level of co-operation went beyond what had previously been deemed possible. In 1990 the signatories enhanced the level of co-operation to phase out all CFCs by 2000, and in 1992 they rolled back the target year to 1996.Footnote ⁶⁰

The United States and Great Britain both played an important role in bargaining during the Protocol negotiations.Footnote ⁶¹ The United States was the world's largest producer of CFCs, accounting for 44 per cent of all CFC production in 1974, followed by Western Europe, which was responsible for about one-third of worldwide production.Footnote ⁶² DuPont (the largest CFC producer in the United States) and Imperial Chemical Industries (ICI), the largest enterprise in Great Britain, were also influential in the negotiation process because their respective countries were particularly attentive to their national firms.

Before the Montreal Protocol, neither DuPont nor ICI had developed alternative chemicals to CFCs. However both companies were interested in the future market profits from developing CFC substitutes because they had ‘the necessary financial and organizational resources to invest in new technologies’ and could ‘build on the experiences of the 1970s, when the first ozone controversy prompted them to look into the potential for substituting CFCs’.Footnote ⁶³ In the late 1970s, DuPont invested about US$70 million into developing substitutes. It revived its research program in 1986, one year before the protocol was signed,Footnote ⁶⁴ and announced that year that CFC substitutes would be available within five years.Footnote ⁶⁵

Although Western European companies, including ICI, did not invest in new substitutes to the same extent, it was clear that ICI also had a comparable capacity to develop new substitutes. For instance, according to patent applications and publications from the mid-1980s, ICI and DuPont both developed synthesis processes and catalysts to produce HFC-134a, one potential alternative to CFC.Footnote ⁶⁶ ICI also resumed a research program into CFC alternatives that it had abandoned in the early 1980s.Footnote ⁶⁷ Due to industrial secrecy, it is difficult to determine how advanced DuPont and ICI technologies were prior to the Montreal Protocol, but we can conclude from the subsequent developments that they had quite symmetric capacities to formulate new substitutes. DuPont announced in 1989 that it would begin commercial production by the end of 1990, and ICI was expected to bring its first HFC-134a facility on stream a few months later.Footnote ⁶⁸

The United States pushed for an international treaty to reduce CFC emissions, but hoped to limit the scope of that treaty to purely non-essential aerosol uses of CFCs. The European Community (EC) was initially reluctant to embrace any international measures to ban CFC emissions, but shifted its position in 1984 by calling for an international production capacity cap and a 30 per cent cut in non-essential aerosol use of CFCs. Each proposal reflected the respective economic interests of the United States and Western Europe: the United States had already begun regulating non-essential uses, and the EC had capped production capacity for CFCs at the 1980 level. Although some measures restricting CFC use were already in place in the United States and Europe, the strongest opposition to creating an international ozone regime came from industry.

Significant progress was made in 1986 when the American industry first acknowledged a connection between CFCs and ozone depletion and when DuPont announced the feasibility of manufacturing substitutes.Footnote ⁶⁹ As the development of CFC substitutes became more feasible, by 1986, industrial representatives significantly softened their opposition against to an international ban on CFCs. The Alliance for Responsible CFC Policy, an industry lobby group, announced that its members were willing to endorse a global ozone regime. DuPont also called for a worldwide limit on CFC emissions.Footnote ⁷⁰ From that point on, the United States assumed a more ambitious position, calling for a long-term reduction of up to 95 per cent of all CFC and halon consumption.Footnote ⁷¹ In addition, the European CommunityEC shifted its position in the summer of 1987 as large European companies recognized the projected market profits from developing new substitutes.Footnote ⁷² A shift in industry position, accompanied with by technologicaly innovation, led the United States and Great Britain to endorse co-operation. As a result, countries were finally able to come to agreement to reduce CFC consumption by 50 per cent.

The successful conclusion of the agreement was due at least in part to the industries’ symmetric ability to develop new substitutes. While the conventional wisdom holds that the American industry was in a superior position, other leading European chemical companies, especially ICI, had sufficient resources and previous experience in developing substitutes. Furthermore, the current scholarly consensus is that ICI came to support CFC regulation because of the expected long-term profits from CFC substitutes.Footnote ⁷³ Davenport also acknowledges that ‘European chemicals industry giants also faced the promise of expanding profits from marketing substitutes’ because ‘new substitutes could be marketed as high-margin specialty chemicals, would thus command higher prices, and ensure a substantial competitive advantage to a few leading international firms that could make that investment’.Footnote ⁷⁴ Jordan writes that Britain's shift in position stemmed from ‘ICI's eagerness not to be excluded from the post-CFC market in substitute chemicals’.Footnote ⁷⁵

Counterfactually, had there been a wide technology gap between DuPont and ICI, Britain would not have ignored ICI's position and would have opposed the signing of the Montreal Protocol. Britain was concerned about ICI's market share (high α_i ) because it was the largest industrial firm as well as the largest CFC producer in Britain.Footnote ⁷⁶ Therefore the British industry's comparable capacity to develop new substitutes was critical for Britain's support. This case shows that the symmetric level of technological development was conducive to international co-operation in the reduction of CFCs. Since American and British industries had similar capacities, major companies expected to benefit from new regulations rather than lose out to competition from rivals.

Neither economic wealth nor environmental interests account for the bargaining outcomes. First, wealth cannot explain why the United States and Britain shifted their negotiating positions in the mid-1980s. Both countries initially provided only limited support for the international ban on CFCs, but they fully endorsed the creation of the ozone regime from the mid-1980s. This shift occurred without any significant change in their economic power. Secondly, the environmental justification also falls short of explaining the timing of the position shift. Sprinz and Vaahtoranta explain why the United States (Great Britain) was more (less) supportive of an international agreement due to its high (low) ecological vulnerability.Footnote ⁷⁷ However, a focus on environmental interests does not explain why Britain changed its position to support the ozone regime during the bargaining process.

Asymmetric Disagreement: The Luxembourg Compromise of 1985

This case study shows that technology competition impeded European co-operation on automobile emission standards. While Germany had developed advanced technology for controlling automobile emissions, Great Britain and France had immature technology. The subsidy cost β_i was low for Germany and high for Great Britain and France. But Great Britain and France still greatly valued market profits (high α_i ) from the automobile industry and refrained from co-operation due to expected losses in a technology competition with Germany.

The debate over automobile emission standards within the EC began in the early 1980s as Germany – together with the Netherlands, Denmark and Greece – pressed for a European bill to adopt stringent limits. Their proposal was opposed by the British, French and Italian governments due to concerns over the technology gap between the German automotive industry and their own. The German industry held advantages on tough emission standards, partly due to tax incentives established by the German federal government. British and French firms lacked such advanced technology, which led them to oppose any European-wide agreement limiting automobile emissions. Due to asymmetries in technology, the European countries at first failed to adopt stringent standards and instead reached a compromise to require three-way catalytic converters to be initially imposed on larger cars with two-liter or greater engines. The EC was only able to slowly expand the scope of co-operation, as British and French car companies attempted to buy the time they needed to develop technology to compete with the German auto industry.Footnote ⁷⁸ Therefore this case illustrates how a technology gap between nations impeded international co-operation.

In the early 1980s, Germany was a leader in engineering technology for controlling automobile emissions. This know-how was unavailable to other national automobile industries in Europe, notably those in Great Britain and France. The German industry had developed an advanced device, the three-way catalytic converter, in order to meet the automotive emission standards of the United States, a key export market for German firms.Footnote ⁷⁹ In addition, the German government's introduction of tax rebates for less-polluting cars had encouraged technological expansion.Footnote ⁸⁰ Wurzel notes that the adoption of tax incentives in Germany played a crucial role in compelling the motor industry to produce vehicles equipped with three-way catalytic converters.Footnote ⁸¹ In constrast, the British and French industries were pursuing an alternative approach, the lean-burn engine technology.Footnote ⁸²

Against this backdrop, Germany proposed stringent emission standards for passenger cars.Footnote ⁸³ However, other nations were concerned about the projected impact on their automobile industries, and generally ascribed a greater weight to these concerns, α_i , than to the expected environmental benefits of the proposal, V_i . Since stringent standards required the adoption of a three-way catalytic converter, key players in the German automotive industry such as Mercedes, BMW and Volkswagen were expected to adapt more easily to a tightening of standards.

Competitors in Britain and France were expected to suffer.Footnote ⁸⁴ The British and French governments expressed their suspicions about Germany's motivations and accused Germany of ‘trying to give its car industry a comparative advantage under a cover of environmental concerns’.Footnote ⁸⁵ In Britain, the German government was viewed as ‘having been persuaded by their manufacturers, on competitive grounds, to press for the catalysts’, and as ‘wanting to achieve a technological hegemony within the automobile sector in Europe’.Footnote ⁸⁶ French car manufacturers also claimed that Germany was ‘saying forest and meaning Mercedes’.Footnote ⁸⁷ Germany threatened to unilaterally require cars sold in Germany to be equipped with three-way catalytic converters.

The fierce British and French resistance stems from the national importance of the car industries (high α_i ) and their technological disadvantage vis-à-vis the Germans. If market profits from the automotive industries had not been important to the British and French economies, they would have co-operated more readily, despite their technological handicap. However, Britain and France had significant interests in automotive market profits. In 1980, road vehicles constituted 8.9 per cent and 8.4 per cent of total exports in Britain and France, respectively.Footnote ⁸⁸ Due to the significant impact of the car industries on their national economies, asymmetric technology magnified their concerns about competition with the German car industry. Also, they argued that the relative costs of adopting the three-way catalytic converter were higher for smaller cars compared to larger cars (on which German manufacturers tended to concentrate), because purchasers of smaller vehicles were more sensitive to price increases than were other consumers.Footnote ⁸⁹

The arduous negotiations at the EC level initially resulted in limited co-operation, the Luxembourg Compromise of 1985, which established distinct emission limit values for three categories of automobiles and required the adoption of a three-way catalytic converter only for large cars with engine displacements over two liters. Emission standards for smaller cars became more stringent, yet remained achievable without a three-way catalytic converter.Footnote ⁹⁰ This outcome reflected the interests of the British and French car manufacturers, which tended to concentrate on the production of small and medium-sized cars. Denmark and Greece vetoed the directive in 1985 because Denmark supported stricter standards for all cars and Greece hoped to receive additional funds for environmental protection. The directive was passed in 1987 following the introduction of the Single European Act, which allowed qualified majority voting.Footnote ⁹¹ This outcome is consistent with our model extension that allows a more enthusiastic country to impose co-operation on others.

Even after the Luxembourg Compromise, the German government continued to support catalytic conversion technology. The federal government introduced tax differentials in favor of unleaded gasoline and low- or reduced-emission vehicles. The government offered tax exemption for anyone registering a new catalytic-converter-equipped car up to DM 3,000 until 1987 and up to DM 1,100 until July 1991. Direct cash subsidies were also offered for retrofitting existing cars with catalytic converters, which operated from January 1986 to July 1992.Footnote ⁹² These measures are consistent with our model's predictions. In 1989, member states agreed to require the adoption of a three-way catalytic converter for small cars under 1.4 liters (the 1989 Small Car Directive) and further expanded the agreement to apply to all new cars by 1992 (the 1990 Consolidated Directive). This path toward broader co-operation faced fewer obstacles because British and French catalytic-converter technologies had grown relatively more symmetrical. Although German manufacturers continued to hold a lead, the British and French industries were now developing proprietary technologies. In fact, the British and French car manufacturers had already begun to develop their technologies prior to the Luxembourg Compromise. For instance, the Financial Times reported in May 1983 that Ford Europe had launched a crash program, investing US$ 200 million and dedicating 400 of its 5,000 European engineers. At the time of the Compromise, however, their technologies were unable to compete with the German industry. The technology gap had narrowed by 1989, and the British and French industries were thus less opposed to new agreements. One British industry representative remarked that they prefer ‘slightly more stringent standards which give us more lead time compared to more lenient standards which stipulate short implementation deadline’.Footnote ⁹³

This case illustrates the impact of technology development on international co-operation. We have argued that the EC's failure to impose stringent emission standards for all cars was due to the asymmetric technology situation and certain member states’ salient interests in market profits from their automotive industries. Had Great Britain and France already possessed more advanced technology analogous to the three-way catalytic converter, the EC would have been able to reach a more profound agreement much earlier. Alternatively, had they held only marginal interests in profits from the automotive industries, the scope of the agreement would have been more expansive. Over time, the EC was able to expand the level of co-operation as the technology gap closed.

As for alternative explanations, the wealth-based explanation fails to account for why wealthy France and Great Britain were hesitant to adopt a European automobile emission standard while economically poorer nations, such as Greece, supported it. Our model can explain this pattern. First, France, Great Britain and Greece were all technology underdogs, but France and Great Britain were highly interested in market profits (high α_i ) while Greece had little interest (low α_i ). Secondly, environmental interests alone are not sufficient to explain this case. While the German public was more concerned about the environment due to the acid rain issue, this view cannot explain why Great Britain and France gradually began to support co-operation, as there was no significant change in terms of environmental interests. Indeed, both Great Britain and France did offer fuel efficiency as an alternative approach to environmental conservation.

Asymmetric Agreement: The European Union Bubble

This case demonstrates the possibility of asymmetric co-operation: European nations were asymmetric in terms of renewable energy technology, but successfully concluded a climate BSA. Denmark, Sweden and Germany possessed advanced renewable energy technology and were leading the industry (high α_i and low β_i ). Other countries, such as Britain and France, were latecomers and had little interest in market profits from renewables (low α_j and high β_j ). European co-operation was possible despite asymmetric technology because technology underdogs had low market valuations.

Under the Kyoto Protocol, the EU agreed to reduce its total greenhouse emissions by at least 8 per cent below 1990 levels by 2012. Before and after the Kyoto conference, EU member states negotiated over how to share the burden of emissions reductions within the union, and reached a burden sharing agreement (BSA) in March 1997 (pre-Kyoto) and again in June 1998 (post-Kyoto). While long-term burden sharing remains, to a certain extent, an ‘unsettled’ issue, this positive negotiation of a BSA is a rare example of success in distributional bargaining within the environmental realm.Footnote ⁹⁴ A key contributing factor for reaching the BSA, as we show, was the ambitious commitments undertaken by countries with a strong interest in renewable energy technology. Denmark and Germany committed to reducing greenhouse emissions by 21 per cent. At the same time, countries such as Portugal and Greece were allowed to increase their emissions by up to 27 and 25 per cent, respectively. Our model makes sense of this outcome by providing a previously unrecognized explanation for the stark asymmetries that characterized the BSA, which are inexplicable from either a purely economic or environmental perspective.

Remarkable variation across member states characterized their varied abilities to deploy renewables. Denmark and Germany led the European renewable energy market in the 1990s. Since the 1970s, the Danish government has promoted wind energy, and to a lesser extent biomass, using a variety of instruments including investment subsidies, public R&D, tax exemptions and a feed-in tariff to private turbine owners.Footnote ⁹⁵ Owing to this support, Danish manufacturers have been the most successful wind turbine producers since the early 1980s. In 1990, Danish wind turbines accounted for some 45 per cent of total worldwide capacity.Footnote ⁹⁶ Germany came to wind power later than Denmark, but it emerged as number two in the world in installed capacity in 1994 and was already number one by 1996.Footnote ⁹⁷ While German industry benefited from the substantial domestic electricity market, government support was also crucial in fostering the solar and wind industries. In 1991, in an attempt to stabilize the market for wind power, the German government passed the Electricity Feed-in Law, which required utilities to purchase renewable electricity and dictated its price.Footnote ⁹⁸ Consistent with our model, (i) public subsidies for renewable energy in Germany and Denmark continued in the implementation stage and (ii) resulted in decreased consumer prices.Footnote ⁹⁹

On the other hand, countries such as Great Britain and France were not particularly interested in renewables. Great Britain had provided R&D support for new energy technologies in the late 1970s, but the government was not pressured to seek alternatives due to abundant hydrocarbon resources and minimal public opposition to nuclear energy.Footnote ¹⁰⁰ Also, the French energy sector was dominated by the nuclear industry, and notably the de facto monopoly Electricité de France, created by President de Gaulle's ambitious nuclear program.Footnote ¹⁰¹ During the Kyoto negotiations, many other countries, from Italy and Belgium to Spain and Portugal, held little commercial interest in renewable energy, as indicated by the data analyzed below. In this context, EU member states understood that some differentiation would be necessary to reach a BSA. The EU had been involved in the global climate negotiations of the early 1990s, but it was not until January 1997 that it introduced a specific proposal for an EU-wide abatement target and individual national goals. The Dutch government suggested a ‘Triptych Approach’, which divided each national economy into three broad sectors: the light domestic sector, the energy-intensive export-oriented sector, and the electricity-generation sector.Footnote ¹⁰² The idea of the Triptych was to develop an objective and economically sound foundation for burden sharing.

The BSA finalized by the EU in 1998 differed from the original Triptych. The commitments, along with other descriptive statistics, are provided in Table 3. Higher per capita numbers of patent applications and larger R&D investment are indicators of a greater ability to commercialize renewable energy. If our hypothesis is correct, countries with high prior per capita patenting rates and R&D investment should accept more ambitious commitments. Using prior patents and R&D allows us to measure each country's interest α_i and cost β_i . High prior patenting and R&D rates suggest both ability and interest, so they map into high α_i and low β_i . In contrast, we cannot use subsequent patents and R&D investment, because our model shows that they are endogenous to co-operation.

Table 3

The European Burden Sharing Scheme, the Triptych and Other Statistics

Note: the data on the number of patent applications filed at the Patent Co-operation Treaty (PCT) and the European Patents Office (EPO) can be acquired from the Organisation for Economic Co-operation and Development Patent Statistics Portal. For patents data, we averaged per capital patent applications for 1993–7. The data on public funding for R&D are available from the International Energy Agency (IEA) database. We only included non-hydropower renewable energy technologies, as defined by the IEA and averaged per capita R&D for 1993–7.

For the test, we fitted a locally smoothed regression (lowess) in the data. First we examined the absolute national commitments. Next we examined deviations from the Triptych. The former approach offers a clean baseline test, whereas the latter approach ensures that the results are robust. The results are reported in Figure 2. Each country's commitment – and the corresponding deviation from the original Triptych – are negatively related to the country's commercial ability in renewable energy. Nations with high patenting rates and large annual investments in R&D into renewables were willing to accept a higher cost for implementing the Kyoto Protocol.

Fig. 2

Locally smoothed regression (LOWESS) results

Note: the upper panels show that the national targets agreed in the EU BSA are negatively related to patents filed at PCT and renewable energy R&D funding. The lower panels show that the deviation from the original Triptych is also negatively related with these measures of prowess in renewables. In each figure, the lines are the non-parametric lowess fits to the data.

The negotiations tell a similar story. Germany was a key player in the negotiation process and showed a willingness to assume an ambitious reduction target from the very beginning. To facilitate the negotiations, in March 1996 Germany proposed a reduction target of 10 per cent by 2005 and 15–20 per cent by 2010 and declared that it would reduce its emissions by 20 per cent. Since Germany's dramatic reductions alone would cover approximately 80 per cent of the EU's overall target, its position played a significant role in the success of negotiations within the EU.Footnote ¹⁰³ In addition to Germany, other ‘rich and green’ countries such as Austria, Denmark and the Netherlands, which had all been investing substantial public funding in renewable energy R&D, contributed to the success of the negotiations by setting their own ambitious targets.Footnote ¹⁰⁴ Technological forerunners expected to benefit from international environmental co-operation, because such efforts would increase demand for renewable energy.

While technological underdogs, such as Greece and Portugal, worried about the cost of emissions reductions, they were not opposed to the agreement as long as they could assume a lighter burden.Footnote ¹⁰⁵ Members of the Cohesion Fund (Greece, Ireland, Portugal and Spain) asserted their right to increase their domestic CO₂ emissions,Footnote ¹⁰⁶ and other members acknowledged the need for differentiated responsibilities.

In sum, technological asymmetry need not cause co-operation failure. If technological leaders are willing to assume a heavier burden, co-operation remains within reach. Germany enthusiastically propelled the negotiation process even at the cost of assuming a considerable burden, and Denmark followed suit. Other countries participated in environmental co-operation because they expected to lose little from the expansion of co-operation. German and Danish enthusiasm would bring about dramatic reductions in the cost of renewable electricity, which would then allow other European countries to comply with the EU BSA at a lower cost.Footnote ¹⁰⁷

Neither wealth nor environmental interests are plausible alternative explanations. First, while economic wealth can explain North-South differences, it cannot explain variation in emission reduction targets among the Northern members. For instance, France and the Netherlands did not deviate significantly from the Triptych, in spite of their wealth. This is also true of Finland and Italy. Secondly, environmental interest cannot explain why some countries assumed a heavier burden than others. According to a 1995 Special Eurobarometer, for example, 82 per cent of surveyed Europeans viewed protecting the environment as an immediate and urgent problem. The variation amongst member states was relatively small and does not capture the different stringencies of nations’ emission reduction targets. For instance, the figure appeared to be highest in Greece (97 per cent) and lowest in Belgium (63 per cent), but Greece assumed a light burden of +25 per cent while Belgium assumed a heavier burden of −7.5 per cent.Footnote ¹⁰⁸