Unpacking Pro-Adjustment Lobbies

We argue that the effect of pro-adjustment lobbies goes beyond simply the balance of pro- and anti-adjustment lobbies. Within the set of pro-adjustment lobbies, there are important differences in the reasons for their support, which condition how lobbies affect the prospects of mutual adjustment (co-operation). The two ‘types’ of lobbies we consider differ by whether they care about (1) policy adjustments made by their government or (2) adjustments made by foreign governments.

The first type of lobby (DBLs) benefits directly from domestic policy decisions: their support for adjustment derives from the policy choices made by their government. The distinguishing feature of these lobbies is their support for their government's adjustment decision in isolation of its effects on other governments’ decisions.

Consider the context of international co-operation regarding CO₂ emissions abatement. DBLs support their government's decision to implement and abide by CO₂ abatement agreements, such as the Kyoto Protocol, because they represent firms that will receive subsidies to develop low-pollution technology. Solar technology firms support CO₂ reduction efforts because subsidies for renewable energy technology are often a prominent part of governments’ abatement plans. They support CO₂ abatement efforts because they benefit directly from domestic policy choices, such as subsidies. Their support for international co-operation to reduce pollution is often not based on altruistic values or an intrinsic desire for lower CO₂ emissions. In the United States, solar firms do not support CO₂ reduction writ large. They have lobbied extensively for import tariffs against Chinese solar technology, because competition from Chinese producers lowers the price of solar cells.Footnote ¹⁸ Such tariffs are contrary to the goal of CO₂ emissions abatement because they raise solar technology prices in the United States, which likely decreases the overall use of the technology.

DBLs are also commonplace in the context of co-operation over lowering trade barriers. In March 2002, the George W. Bush administration took a significant step away from its overall support of free trade and placed heavy tariffs on steel imports from major producers such as the European Union. The move was largely political, designed to increase support from politically important domestic steel producers in the 2002 elections that November. In November 2003, the World Trade Organization (WTO) ruled against the legality of the tariffs, and the Bush administration had to choose whether to re-adjust its policy back toward free trade or keep the tariffs in place. While the administration could abide by the WTO ruling and remove the tariffs, it could also ignore the ruling and leave them in place, since the tariffs had not yet reached their predetermined expiration date. Despite the WTO ruling, US steel producers urged the administration to leave the tariffs in place.

Many lobbies, however, advocated compliance with the WTO ruling, and the steel tariffs were lifted in December 2002. Many of these anti-tariff groups can be thought of as DBLs: they opposed the tariffs because they benefitted directly from adjustment. The steel tariffs had a larger-than-expected effect on domestic prices, and the price of steel increased substantially. Producers that used steel as an intermediate good suffered; thus politically powerful interest groups, such as those representing American automobile manufacturers, strongly supported removing the steel tariffs.Footnote ¹⁹ These lobbies supported compliance because they benefitted directly from lower prices for production inputs.

In contrast to DBLs, which benefit from their own government's policies, the second type of lobby (IBLs) benefits from how their government's adjustment decision affects the actions of other governments. Firms and other actors in one country can often benefit from, or be harmed by, the policy adjustments made in another country. In today's globalized world, the policy decisions of foreign governments can have significant impacts on the welfare of domestic firms. IBLs support their government's decision to adjust because they know that the decisions of their government and its foreign partners are related. Under reciprocity, where a foreign government adjusts if and only if the other government adjusts, lobbies that care about foreign governments’ decisions also care about their own government's decisions.

Again consider the context of CO₂ abatement. While firms installing solar panels at home are DBLs, solar technology firms that expect to export their products to other countries are IBLs; they benefit when those other countries implement policies that encourage demand for solar cells, like rebates or tax breaks for consumers who choose this technology. A domestic solar producer might expect to benefit from a foreign trading partner's decision to engage in CO₂ abatement efforts. Such firms might support their own government's pollution abatement efforts, since the policy adjustments of foreign countries might be contingent on their government's decisions.

These types of lobbies have also been emphasized in explanations of international trade co-operation. Groups representing the interests of exporting firms often oppose tariff barriers on goods in unrelated sectors or industries, because those tariffs can provoke harmful retaliation from the governments of trade partners.Footnote ²⁰ Returning again to the steel tariffs example, many relevant interest groups were IBLs: they opposed the tariffs because they feared retaliation from the United States’ trading partners. In response to the steel tariffs, the European Union promised retaliatory tariffs on imports of US products that were produced by exporting firms located in politically sensitive states.Footnote ²¹ Textile producers in the Carolinas and orange growers in California strongly supported compliance with the WTO ruling because they feared the consequences of this retaliation.

In practice, lobbies’ preferences can have both domestic and international elements: a lobby's welfare can be affected by both domestic and foreign policy decisions. Some lobbies care about what their government does and what foreign governments do, so the domestic-international difference is not black and white. We could also imagine lobbies that care about both domestic and international actions, but that care more about one or the other.Footnote ²² We describe two ‘types’ in order to highlight this important dimension of variation in lobby preferences.

Variation in pro-adjustment lobbies is important because international co-operation often resembles a repeated prisoner's dilemma. Governments would like to mutually choose adjustment, but have incentives to defect from co-operation unilaterally. The shadow of the future, and the possibility that unilateral defections would be met with punishment, help sustain co-operation. As the model below demonstrates, the type of pro-adjustment lobby under consideration and the intensity of its preferences can affect the credibility of punishment strategies, and in turn, the sustainability of co-operation.

The Game Without Lobbies

In each period of an infinitely repeated game, governments i and j simultaneously choose between adjustment and defection; that is, the choice not to make a particular policy adjustment. If government i adjusts, then government j receives a benefit, b_j > 0, and if government j adjusts, then government i receives the benefit b_i > 0. A government choosing to adjust pays a cost, c > 0.Footnote ²⁹ This cost can be thought of as a combination of the economic, administrative and political costs of the policy adjustments required for co-operation. Notably, the cost term also represents the political cost to the government of imposing costs on anti-co-operation lobbies, such as import competitors in the case of trade. In the interest of simplicity, we do not endogenize the behavior of these lobbies. However, their effects on outcomes can be evaluated by varying the cost c.

Mutual adjustment is assumed to be optimal for global welfare: $$$ {{b}_i},{{b}_j}\, \gt \,c $$$ . All players discount future pay-offs by some common factor $$$\delta \in ({\rm{0,1}}) $$$ . We assume that the governments are identical to one another, aside from their benefits to adjustment, so that we can isolate the effects of variation in their domestic lobbies.

The Game With Lobbies

In games with lobbies, governments play the same game described above, but each also interacts with a lobbying group within its country, lobby i and lobby j. Lobbies also differ from one another in two additional ways. First, they can differ by whether they gain benefits when their government adjusts (DBLs) or when the foreign government adjusts (IBLs). An IBL in country i gains a benefit V_i > 0 when the foreign government chooses to adjust. In other words, when government j adjusts, lobby i gains V_i , and when government i adjusts, lobby j gains V_j (and vice versa for an IBL in country j). A DBL in country i gains a benefit V_i > 0 when its government chooses to adjust: when government i adjusts, lobby i gains V_i , and when government j adjusts, lobby j gains V_j (and vice versa for a DBL in country j).

Secondly, lobbies can vary in the intensity of their preferences. We do not assume that V_i = V_j for any of the lobby-type permutations considered below. This allows us to consider situations in which one lobby might value adjustment more or less intensely than its foreign counterpart. Also note that we do not assume that the lobbies’ pay-offs for adjustment are identical to their governments’ pay-offs. Allowing lobby and government preferences to diverge, and allowing lobby preferences to diverge from one another, lets us consider a rich variety of preference constellations.

In the game with lobbies, each of the two lobbies, i and j, can influence its own government's decision by offering a reward to its government when it chooses to adjust. The rewards, denoted r_i and r_j , can be thought of as conventional campaign contributions or other direct lobby influence, as in Grossman and Helpman.Footnote ³⁰ Within each period of the repeated game with lobbies, the sequence of the game is as follows: first, the lobbies simultaneously select the reward to offer their government for its choice to adjust, with each reward greater than or equal to zero.Footnote ³¹ Secondly, the governments observe the rewards offered by the two lobbies and simultaneously choose to adjust or defect. If government i adjusts, the reward, r_i is added to its pay-off and subtracted from lobby i's payoff. Similarly, if government j adjusts, the reward, r_j , is added to its pay-off and subtracted from lobby j's pay-off.Footnote ³²

Our interest is in the range of parameter values that allows governments to co-operate in a subgame perfect Nash equilibrium using grim trigger punishment. However, we must also specify equilibrium strategies for the lobbies. We assume the lobbies play their one-shot optimal actions in each period. That is, a lobby's action in each period must be a best response to other players’ actions during that period. In other words, lobbies are not allowed to use contingent strategies. For example, lobby i cannot agree to reward government i as long as lobby j rewards government j.

An advantage of this approach is that each lobby's punishment strategy is individually rational by definition. This means that the credibility of the grim trigger punishment, from a lobby's perspective, can be easily characterized. The limitation is that we cannot analyze lobbies’ ability to improve their pay-offs through contractual arrangements akin to international treaties. This simplifies the analysis and allows us to focus on the incentives of governments. Moreover, it could be difficult for lobbies to write binding contracts with foreign governments or lobbies. Therefore, our simple approach is a logical starting point. In the conclusion, we expand on lobbies’ strategic incentives to change their behavior due to the possibility that they will spoil co-operation.

Without Lobbies

In the absence of lobbies, it is clearly not possible to sustain co-operation without repeated play, since neither government obtains a benefit from unilateral adjustment. The unique Nash equilibrium in any single period is for both governments to defect. We first establish the conditions under which a grim trigger punishment strategy can sustain co-operation on equilibrium path in repeated play.

1. 1. Both governments adjust in the first period.

2. 2. If both governments adjust in period t, then they also adjust in period t + 1.

3. 3. If either government defects in period t, both governments defect forever beginning in period t + 1.

In the repeated game without lobbies, co-operation is sustainable under familiar reciprocity conditions: (1) when the governments’ discounted benefits from co-operation are high enough relative to their costs and (2) when the governments have a high enough discount factor. If the condition established in Claim 1 is met, so that δb_i , δb_j ≥ c, the threat of grim trigger is sufficient to ensure that both governments want to adjust. This punishment threat is credible, because in subgames in which either government has previously defected, neither has any incentive to adjust, knowing that the other government will defect. Without lobbies, co-operation is possible when the discounted value of co-operation exceeds the cost. Figure 1 shows the range of benefits for the governments in which co-operation is sustainable. Each government must ascribe a sufficiently high value, relative to the cost and adjusted for time preferences, to the other government's adjustment.

Fig. 1

Regions of co-operation: no lobbies

Equilibrium Analysis: IBLs

In the sections that follow, we investigate how lobbies can enhance the possibility of co-operation even when governments are unable to sustain co-operation on their own. Specifically, for analyzing the games with lobbies, we assume that the conditions established in Claim 1 are not met. In other words, co-operation is impossible under the specified punishment strategies without lobby influence. This allows us to demonstrate the conditions under which lobbies that care about adjustment (and can condition rewards for their government based on its choices) are able to facilitate co-operation where it would otherwise have been impossible based on the two governments’ preferences alone.

We now consider conditions for co-operation when both countries have IBLs that value adjustment by the foreign government. We consider an equilibrium in which (1) lobbies can offer non-negative rewards to their government when it chooses to adjust and (2) lobbies and governments punish each other for defections with lowered contributions and future defections, as in the grim trigger punishment strategy above. We use superscripts to denote the reward offered in a particular period (for example, the reward offered by lobby i during period t is $$$r_{i}^{t} $$$ ).

1. 1. In period 1, the lobbies offer $$$ r_{i}^{1} \, = \,r_{i}^{\ast} \, = \,c\,{\rm{ - }}\,\delta \,{{b}_i} $$$ and $$$r_{j}^{1} \, = \,r_{j}^{\ast} \, = \,c\,{\rm{ - }}\,\delta \,{{b}_j}$$$ . Government i adjusts if $$$r_{i}^{1} \geq r_{i}^{\ast} $$$ and government j adjusts if $$$ r_{j}^{1} \geq r_{j}^{\ast} $$$ .

2. 2. If both governments adjusted in period t, then government i adjusts in period t + 1 whenever $$$ r_{i}^{{t + 1}} \geq r_{i}^{\ast} $$$ and government j adjusts in period t + 1 whenever $$$ r_{j}^{{t + 1}} \geq r_{j}^{\ast} $$$ .

3. 3. If both governments adjusted in period t, then lobby i offers $$$r_{i}^{{t + 1}} \, = \,r_{i}^{\ast} $$$ and lobby j offers $$$r_{i}^{{t + 1}} \, = \,r_{j}^{\ast} $$$ in period t + 1.

4. 4. If either government defects in period t, then for every period k > t, lobbies offer $$$ {{r}^k} \, = \,0 $$$ and government i adjusts if and only if $$$ r_{i}^{k} \geq c $$$ and government j adjusts if and only if $$$r_{j}^{k} \geq c$$$ .

Claim 2 holds that the presence of IBLs can facilitate co-operation even where it would otherwise have been impossible in the absence of lobbies. The condition in Claim 2 says that lobbies must value adjustment enough to want to offer (and pay for) these rewards in each period. In this equilibrium, the lobbies offer their governments positive rewards, $$$ {{r}^\ast} $$$ , which induce their government to adjust. This equilibrium exists when the lobbies’ values for adjustment, V, are higher than the smallest possible reward they must offer their government to induce adjustment.Footnote ³³

The smallest possible reward needed to induce adjustment is determined by the difference between their government's pay-off stream with co-operation and their temptation to defect. The present value of the government's net-of-rewards utility stream to adjustment is $$$\frac{{b\,{\rm{ - }}\,c}}{{1\,{\rm{ - }}\,\delta }}$$$ and the government would get a pay-off of b if it unilaterally defected. The lobbies offer just enough of a reward to ensure that their government's pay-off for continuing to adjust, which is $$$ \frac{{b\,{\rm{ - }}\,c\, + \,r}}{{1\,{\rm{ - }}\,\delta }} $$$ , is equal to their government's pay-off for defecting. In other words, the lobby offers just enough to keep their government from defecting, and no more, since any additional reward is more costly for the lobby yet gains them no additional benefit.

If a government defects, either of its own volition or because its lobby offered it an insufficiently high reward, then the lobbies will withdraw their future rewards and the governments will defect in the future. As before, a government's threats to withdraw from co-operation are credible because it gains nothing from adjustment, knowing that the other government's defection will deny it any benefit to adjustment and that its lobby is no longer offering rewards for adjustment.

IBLs’ threats to withdraw their rewards are also credible. IBLs only care about the foreign government's decision to adjust or defect. After a government has defected, the lobby in country i cannot influence the decision of government j, and vice versa. The lobby can do no better by offering a positive reward, because the reward to the home government will not change the foreign government's decision and would be costly if their government accepts the reward and adjusts.

Since the lobbies are willing to compensate their governments for adjustment, in order to avoid the grim trigger, they enable additional co-operation. Figure 2 illustrates the regions in which the presence of IBLs can facilitate co-operation. In the cross-hatched regions, both lobbies value adjustment enough to make their optimal (positive) reward offers to their government. In these regions, their governments co-operate on sn equilibrium path, even though co-operation would be impossible without the lobbies.

Fig. 2

Regions of Co-operation: Internationalist Lobbies

Some of the well-known examples from the context of international trade described above demonstrate this concept. MilnerFootnote ³⁴ has argued that exporters were able to pressure their governments to resist otherwise tempting protectionist pressures because the exporters wanted to avoid retaliatory tariffs by partner governments. More recently, IBLs have influenced the economic relations between the United States and the rapidly growing China. In the United States, Congress has repeatedly threatened to impose sanctions against China unless the Chinese government reduces the extent of its alleged undervaluation of the yuan. While such legislation may appease domestic import competitors, it scores few points among exporting interests. On 18 October 2011, Thomas Donohue, the president of the US Chamber of Commerce, America's premier business lobby, warned Congress of the negative consequences of punishing China through protectionist legislation: ‘They [China] want to keep all those people working and if [currency appreciation] was forced on them somehow … they would simply just drop the prices as low as they have to to keep those folks working’.Footnote ³⁵

In the vocabulary of our model, the US Chamber of Commerce is an IBL because it worries about foreign retaliation against domestic economic policies. It fears that China would respond to unilateral economic sanctions by adopting policies that induce further harmful economic distortions. Groups such as the Chamber of Commerce can potentially increase the United States’ ability to engage in economic co-operation with other countries. When lobbies are as influential as the Chamber of Commerce, there is little doubt that legislators in Washington will respond to their concerns.

Equilibrium Analysis: Domestically Benefiting Lobbies

We now consider conditions for co-operation in a dyad in which both countries have DBLs that value adjustment by their own government. To maintain comparability with the previous games, we consider an equilibrium in which (1) lobbies can offer non-negative rewards to their government when it chooses to adjust and (2) lobbies and governments punish each other for defections with lowered contributions and future defections, as in the grim trigger punishment strategy above. This equilibrium is characterized by co-operation with the threat of punishment.

1. 1. In period 1, the lobbies offer $$$r_{i}^{1} \, = \,r_{i}^{\ast} \, = \,c\,{\rm{ - }}\,\delta \,{{b}_i}$$$ and $$$ r_{j}^{1} \, = \,r_{j}^{\ast} \, = \,c\,{\rm{ - }}\,\delta \,{{b}_j} $$$ . Government i adjusts if $$$r_{i}^{1} \geq r_{i}^{\ast} $$$ and government j adjusts if $$$r_{j}^{1} \geq r_{j}^{\ast} $$$ .

2. 2. If both governments adjusted in period t, then government i adjusts in period t + 1 whenever $$$r_{i}^{{t + 1}} \geq r_{i}^{\ast} $$$ and government j adjusts in period t + 1 whenever $$$r_{j}^{{t + 1}} \geq r_{j}^{\ast} $$$ .

3. 3. If both governments adjusted in period t, then lobby i offers $$$r_{i}^{{t + 1}} \, = \,r_{i}^{\ast} $$$ and lobby j offers $$$r_{i}^{{t + 1}} \, = \,r_{j}^{\ast} $$$ in period t + 1.

4. 4. If either government defects in period t, then for every period k > t, lobbies offer $$$ {{r}^k} \, = \,0 $$$ and government i adjusts if and only if $$$r_{i}^{k} \geq c$$$ and government j adjusts if and only if $$$r_{j}^{k} \geq c$$$ .

Note that the behavior specified in this equilibrium is identical to that analyzed in Claim 2, except for the conditions under which these strategies constitute an equilibrium. DBLs present an additional complexity, due to the possibility that a lobby would want to unilaterally induce its government to adjust, regardless of the adjustment decision of the other country. Given this problem, we also characterize a second equilibrium of this game, which results in policy adjustments by both governments without conditional punishments (that is, harmony).

1. 1. In every period t, the lobbies offer $$$r_{i}^{t} \, = \,r_{j}^{t} \, = \,r_{i}^{\ast} \, = \,r_{j}^{\ast} \, = \,c$$$ .

2. 2. Government i adjusts if and only if $$$r_{i}^{t} \geq c$$$ and government j adjusts if and only if $$$r_{j}^{t} \geq c$$$ .

Figure 3 shows the two regions in which mutual adjustment occurs, which correspond to these two equilibria. Claim 3 describes the co-operation that occurs in the smaller, middle box. Claim 4 describes the harmony that occurs in the larger box in the top right-hand corner. Like IBLs, DBLs can help facilitate co-operation relative to the game without lobbies. However, the figure shows a key difference between the prospects of co-operation for IBLs and DBLs. With DBLs, the range of parameter values that supports unilateral adjustment, in which at least one government adjusts, is larger. But the range of parameter values that supports mutual adjustment is smaller.

Fig. 3

Regions of co-operation: domestic lobbies Note: The middle cross-hatched box corresponds to co-operation with punishment (Claim 3). The upper-right cross-hatched box corresponds to adjustments without punishment (Claim 4).

This key difference arises because DBLs that value adjustment more strongly than their foreign counterparts suffer from an inability to commit to punishing their own governments. As a result, they make it harder for the foreign lobby to induce its own government to adjust. Consider what happens when lobby i's value to adjustment by government i is greater than the costs to adjustment: $$${{V}_i}\geq c$$$ . In any period, lobby i is willing to offer a reward that completely offsets the government's costs of adjustment: $$$ {{r}_i}\, = \,c\, + \,{\epsilon} $$$ , regardless of the strategies of lobby j and government j.Footnote ³⁶ Government i is willing to accept this offer and choose to adjust. If lobby i values adjustment enough, then it cannot credibly commit to withholding rewards from government i, even when government j defects.

The intensity of lobby i's preferences, in turn, has a subtle but important effect on lobby j's strategy. Lobby i's inability to commit to withholding rewards as punishment increases the reward that lobby j must offer government j for adjustment. Recall that lobbies want to offer the smallest possible reward that offsets the difference between their government's costs and the benefits of adjustment. When lobby i cannot commit to withholding rewards as punishment, the costs of defection by government j decrease. When government j defects, this no longer causes lobby i to withhold rewards and government i to defect, which makes defection relatively more attractive for government j. Defection by j no longer entails reciprocal defections by i. Lobby j must now offer a higher reward to government j in order to offset this increased difference between j's pay-offs from defection versus adjustment. Specifically, lobby j must also offer a reward that completely offsets government j's costs of adjustment, $$$ {{r}_j}\, = \,c $$$ , since government j bears no additional ‘punishment costs’ for defecting. If government i adjusts regardless of government j's decision, then it becomes more expensive for lobby j to induce its government to adjust.

If one lobby values adjustment strongly enough to offer adjustment-guaranteeing rewards, but the other lobby does not, then the types of strategies considered above do not facilitate co-operation in equilibrium. The asymmetry between the lobbies undermines reciprocity by removing an important cost to free riding: punishment by co-operative partners. Proposition 2 summarizes the relationships between domestic lobbies and the prospects for co-operation.

A comparison of Propositions 1 and 2 is illustrative. With IBLs, co-operation is easier compared to a world without lobbies. IBLs value adjustment, so they are willing to help defray their government's costs and make adjustment relatively cheaper. However, they do not compromise the credibility of punishment. Since they have no incentive to influence their own government once reciprocity has failed, these lobbies do not undermine the credibility of contingent adjustment. To sustain co-operation through reciprocity, as KeohaneFootnote ³⁷ has proposed, IBLs are ideal.

DBLs can also make co-operation easier than a world without lobbies, but they are less effective when their preferences are too intense. While they give governments additional incentives to adjust, they also have a downside, compared to the world without lobbies. Since DBLs want their government to adjust even unilaterally, these lobbies can also undermine reciprocity. A government with DBLs that are willing to pay a lot for unilateral action cannot threaten foreign governments with suspending adjustment, for this threat would not be credible, making defecting a more attractive option.

While there are several obstacles to credible climate co-operation, the fact that many environmental groups want their own government to act unilaterally is one of them. Especially in Europe, environmental groups have been successful in inducing the European Union to mitigate climate change regardless of what other major emitters do. Given this interest in unilateral action, these groups’ political clout means that Europe's commitment to climate mitigation is not conditional on reciprocal mitigation by other major emitters, like China or the United States. Our model suggests that these groups’ activities could cause the replacement of contingent co-operation with a pattern of unilateral action on the part of a self-proclaimed ‘climate leader’, such as the European Union. If Europe could threaten other major emitters with suspension of climate policy if there is no progress in international negotiations, other major emitters would understand that any progress in climate mitigation also depends on their policies. In other words, the other major emitters could not free ride on Europe's efforts to mitigate climate change and promote clean energy.

Testing this claim empiricallyis difficult, because rational governments generally would not negotiate a treaty that is expected to fail.Footnote ³⁸ This is probably one of the reasons why extant climate treaties, including the Kyoto Protocol, are not very ambitious and do not contain credible enforcement mechanisms.Footnote ³⁹ However, one may examine the validity of the theoretical result by examining lobbies’ activities in climate negotiations. If lobbies’ activities focus on their own government's policy, calling for unilateral action, and if the negotiations fail because major emitters do not trust each other's promises to co-operate, then the empirical record is consistent with the model's predictions.Footnote ⁴⁰

This problem with the environmentalists’ strategies has complicated multilateral climate negotiations. During the year preceding the infamous Copenhagen climate summit in December 2009, which was expected to lay the foundation for a global climate treaty, many environmental groups lobbied the European Union for unilateral leadership. On 27 January 2009, Joris Den Blanken and Sebastien Risso of Greenpeace's European Union Unit (GPEUU) called on the EU ‘to reduce emissions by at least 40 per cent in 2020, compared to 1990 levels’ without any mention of conditionality.Footnote ⁴¹ On 28 January 2010, when Copenhagen failed to produce a global treaty following dramatic turns in the negotiations, GPEUU criticized the European Union for failure to show more unilateral leadership. As Den Blanken then explained:

Again, Greenpeace argued that the European Union should increase the ambitions of its climate policy regardless of how China, the United States and other major emitters behave. According to our model, which provides firm microfoundations for Victor'sFootnote ⁴² conjecture, this strategy is counterproductive for sustaining reciprocal climate co-operation: ‘Doing better will require that climate diplomats push countries to make contingent offers and realize that the form of contingency will vary with the interests and abilities of governments.’Footnote ⁴³

The model suggests that the domestic power balance between IBLs and DBLs is important for the credibility of reciprocal adjustment. In the climate case, the validity of this conjecture could be scrutinized by evaluating governments’ credibility in different political-economic contexts. If environmental groups that call for unilateral action receive support from industries that prefer unilateral action, such as solar panel installers, we expect the government's credibility to be limited. However, if industries with a strong preference for the foreign government's action (such as exporters of clean energy technologies) are powerful, then the government's credibility should be bolstered. Based on the history of climate co-operation, one would expect non-credible promises to be a more severe problem for Europe, where environmental groups insist on unilateral action and have considerable political clout, than for China and the United States, where these interests are politically less powerful relative to the heavy industry and fossil fuel producers.

Equilibrium Analysis: Mixed Dyads

Finally, we consider the conditions for co-operation in a ‘mixed’ dyad in which lobby i is an IBL and lobby j is a DBL. In other words, both lobbies gain value from the adjustment decision of government j. We again consider the types of trigger strategies described above in which lobbies can (potentially) condition rewards and governments can condition co-operation on past adjustment.

1. 1. In period 1, the lobbies offer $$$r_{i}^{1} \, = \,r_{i}^{\ast} \, = \,c - \delta {{b}_i}$$$ and $$$r_{j}^{1} \, = \,r_{j}^{\ast} \, = \,c - \delta {{b}_j}$$$ .

2. 2. If both governments adjusted in period t, then government i adjusts in t + 1 if $$$r_{i}^{{t + 1}} \geq r_{i}^{\ast} $$$ and government j adjusts in t + 1 if $$$r_{j}^{{t + 1}} \geq r_{j}^{\ast} $$$ .

3. 3. If both governments adjusted in period t, then lobby i offers $$$r_{i}^{{t + 1}} \, = \,r_{i}^{\ast} $$$ and lobby j offers $$$r_{j}^{{t + 1}} \, = \,r_{j}^{\ast} $$$ .

4. 4. If either government defects in period t, thenfor every period k > t, lobbies offer $$$ {{r}^k} \, = \,0 $$$ and government i adjusts if and only if $$$ r_{i}^{k} \geq c $$$ and government j adjusts if and only if $$$ r_{j}^{k} \geq c $$$ .

In a mixed dyad, co-operation is again not sustained in equilibrium with grim trigger-type punishment strategies when lobby j values government j's adjustment strongly enough to unilaterally fund adjustment, regardless of government i's decision. Since government i and lobby i's benefits come from adjustment by government j, they have no incentive to adjust when lobby j unilaterally funds adjustment by government j. Lobby j sufficiently rewards government j to induce adjustment, while lobby i avoids paying any rewards and government i avoids paying any costs for its own adjustment. Lobby j's inability to commit to refraining from rewarding adjustment by government j leaves lobby i with no incentive to offer a reward for government i to adjust.

It is important to note that in this case, both the IBL and its government obtain their maximal pay-offs. They need not contribute or co-operate, and yet the DBL induces its government to adjust. Thus co-operation fails because the game's outcome is actually in the collective interest of one of the countries.

Co-operation is still possible in the region where lobby j has an ‘intermediate’ value to government j's co-operation. Having these intermediate preferences means that lobby j can still credibly commit to withdraw its rewards if government i no longer adjusts. Lobby j likes adjustment enough to offer a reward on an equilibrium path, but not enough to offer a sufficiently strong reward to induce adjustment off the equilibrium path. Figure 4 shows the region in which co-operation is possible: it is identical to that of a dyad with two internationalist lobbies, except for the removal of the region where $$${{V}_j}\, \gt \,c$$$ .

Fig. 4

Regions of co-operation: mixed dyad

Feasible Equilibria with Alternative Punishment Strategies

So far, we have examined the conditions under which different types of lobbies facilitated or hindered co-operation under grim trigger-type punishment strategies. A natural question is how these findings are affected by the consideration of alternate punishment schemes. Specifically, if one DBL valued adjustment much more strongly than its foreign counterpart, and co-operation failed, is there a punishment scheme that sustains mutual adjustment? In general, the answer is yes. The punishment schemes described in Folk Theorem arguments can sustain co-operation if the players are sufficiently patient – that is, for a high enough discount factor.

Consider the possibility that the common discount factor δ has a very high value, δ → 1. In this case, it is possible to construct a punishment strategy that sustains co-operation.

The details of the punishment strategies that sustain co-operation in Proposition 3 are are adapted from Fudenberg and Maskin and are provided in the Appendix.Footnote ⁴⁴ Intuitively, the punishment strategy is based on the idea that even small punishments can have large pay-off effects. Since even slight changes in the rules of co-operation can deter governments and lobbies from defecting, the severe grim trigger punishment is not needed.

Unfortunately for proponents of co-operation, these punishment schemes require that governments discount future pay-offs only a little. Their success depends on the imposition of slight costs on deviating governments over long periods of time. If governments are interested in their short- and medium-term pay-offs, the elaborate strategies detailed in the Appendix do not sustain co-operation. This means that the practical relevance of these strategies is probably limited.