Battle of the Sexes Games and Rambo Situations Within the ExtraRegional Logic

Within the extra-regional logic of regional integration, the structure of the games member states play against each other depends on the reactions of extra-regional actors. On the one hand, extra-regional actors may systematically reward regional integration. This means that investments increase only for cooperative member states of regional organisations and that preferential market access is only granted to these member states as well. If single member states did not implement regional rules, they would suffer from declining extra-regional investment inflows and losses of access to extra-regional markets. When extra-regional actors support regional integration in this way, the member states of the respective regional organisations find themselves in battles of the sexes (see the left-hand table in Fig. 2.3). Although collective action in respect to common pool resources is usually associated with prisoner’s dilemmas (Hardin 1968), the situation within the extra-regional logic is different because regional member states do not try to avoid an unsustainable exploitation of such resources, but rather aim to improve the access of their region to these resources. Free riding is not a problem for the regional member states as long as regional defection is sanctioned systematically by extra-regional actors through declining investment inflows and a denial of market access. The regional member states may nevertheless have different preferences about common regional rules, because different rules may have different distributive consequences. For example, regional member states may disagree about rules of origin when establishing free trade areas or about common external tariffs when establishing customs unions. Thus, member states need

Fig. 2.3 Battle of the sexes and the Rambo situation within the extra-regional logic

to agree to one cooperative solution among many, but once agreements are reached, implementation needs to be enforced by extra-regional actors who sanction regional defection with declining investment flows or the suspension of preferential market access.

On the other hand, the reaction of extra-regional actors may also be less systematic and less in favour of regional integration, which moves the structure of the game from a battle of the sexes towards a Rambo situation.¹ Such Rambo situations are asymmetrical games, wherein one player—unlike the other(s)—lacks any incentive to cooperate, and can only lose out if an agreement is reached. This Rambo’s dominant strategy is therefore to defect, while the other player(s) have the dominant strategy to cooperate (see the right-hand table in Fig. 2.3; for Rambo situations in game theory, see Holzinger 2003; Zurn 1992, 1993). In contrast to prisoner’s dilemmas, tit-for-tat cannot produce cooperative solutions in Rambo situations because Rambos do not defect in order to free ride from the cooperation of others, but because they have lost any interest in cooperation at all. Thus, the other players of the game cannot effectively punish Rambos by answering with defection themselves.

Rambo situations occur whenever one of the regional member states enjoys privileges in its economic relations to important extra-regional actors. It may well happen that certain member states are more attractive as economic partners for extra-regional actors than are their neighbours within the same region. This may be motivated by various economic, political or cultural reasons. For example, due to market size, investments in some member states may be more attractive than investments in others;

some member states of regional organisations may be more important as part of the security strategies of extra-regional actors; and some regional member states may have more cultural ties to certain extra-regional actors because they are former colonies. The result of such circumstances is that the member states that enjoy special attention from other world regions become privileged within their own regions. Such privileged positions constitute competitive advantages for the respective member states in comparison to their regional neighbours, including disproportionally high shares of extra-regional investment and export flows to the disadvantage of their neighbours.

The extra-regional privileges of certain member states may be at odds with regional integration if regional cooperation requires giving up these privileges. For example, privileged member states need to give up bilateral trade agreements with extra-regional partners when the establishment of customs unions requires the harmonisation of external trade regimes. In such circumstances, the privileged member states have to calculate what counts more: their own share of the intraregional and extra-regional gains of regional integration or the losses of their extra-regional privileges. As long as their shares of the intraregional and extra-regional gains from regional integration exceed the losses of privileges, the games remain battles of the sexes. However, if the losses of privileges weigh more than the gains of regional integration for certain member states, these states become Rambos with a dominant strategy of defection in order to protect their privileges.

If Rambo situations occur before the adoption of certain regional agreements, they lead to deadlocks in decision-making that cannot easily be overcome with the help of regional institutions. In battles of the sexes, regional institutions that offer majority rule, agenda setting or even full delegation of decision-making may help to overcome such deadlocks because all member states prefer any cooperative solution to non-cooperation. Thus, they accept giving up sovereignty in order to not endanger the common project. However, such common interest in cooperation does not exist in Rambo situations, wherein at least one member state has no interest in cooperation at all. Such Rambos cannot accept majority rule or delegation of decision-making because these would lead to cooperative solutions, which reduce the Rambos’ pay-offs.

In the case of Rambo situations, deadlocks can only be solved by granting side payments to the defecting member states or by packaging large deals, which combine several cooperation problems with inversed-preference constellations. However, side payments and package deals cannot effectively be facilitated by establishing the usual regional institutions of majority rule or delegation of decision-making competencies; they require high-level intergovernmental negotiations. Side payments always mean the redistribution of wealth from one member state to another, which implies that they need high degrees of political support, and this support can only be provided by the governments of the member states. Package deals require that negotiators overlook a large number of cooperation problems, and are able to package them into one deal. As a result, the specialisation of decision-making into different committees or working groups does not make much sense, because package deals must necessarily transcend the borders of particular issue areas. In contrast, one can expect that summit diplomacy plays a very important role for negotiating side payments and package deals. Only at the top political level are negotiators able to provide the necessary political support for side payments or to combine several issue areas into one package deal. This importance of high-level negotiations for regional integration is grasped under the term ‘interpresidential- ism’ in the academic literature (Malamud 2003, 2005).

If Rambo situations occur after the adoption of certain regional agreements, defection takes the form of non-implementation of regional rules. In such cases, the member states have already agreed to regional rules, but changing circumstances lead to some member states profiting from defection and they become regional Rambos. Due to the inertia of regional institutions (Pierson 2000; Scharpf 1988), it is unlikely that Rambos will be able to turn back the integration process and to renegotiate regional agreements. However, regional Rambos are always able to undermine integration processes by not implementing regional rules that stand against their dominant interests. Thus, regional institutions still exist formally, but their actual effects are limited. If this happens repeatedly, a decoupling of formal institutions and real effects occurs. In the very end, regional integration may even become a Potemkin village with little or no influence on the real economy of the member states.

Legalisation is unlikely to improve implementation in the face of Rambo situations because sanctions by other member states cannot effectively reinforce judicial rulings. Legalisation facilitates cooperation in iterated prisoner’s dilemmas because it stabilises the tit-for-tat strategies of the member states. Such tit-for-tat means that defection of one member state is sanctioned by the defection of the other member states in the following round of the game (Axelrod 1984). Here, dispute settlement mechanisms determine which member states are violating regional rules and can thus be sanctioned legitimately by other member states. However, member states can only sanction each other when they are able to hurt single free riders and when all states have an overall interest in regional cooperation. Without the possibility of exclusion, the member states cannot prevent free riders from consuming the collective goods. In addition, if Rambos do not have any interest in cooperation at all, exclusion from the regional group does not harm them. When the competition for the common pool resources extra-regional investments and exports leads to Rambo situations, both the possibility of exclusion and the overall interest in cooperation are not given. Although Rambos can be excluded from regional organisations, they cannot be excluded from extra-regional investment and export flows, because it is extra-regional actors, and not the regional organisations, that decide on the distribution of these goods. Furthermore, Rambos do not try to free ride on the cooperation of others, but instead they have no interest in regional cooperation at all. Thus, judgements of regional dispute settlement mechanisms cannot be effectively reinforced by decentralised sanctions of member states.

To sum up, within the extra-regional logic member states either face battles of the sexes or Rambo situations—depending on whether extraregional actors reward regional integration systematically or grant extraregional privileges for some member states. Although battles of the sexes are problematic situations because member states have to agree on one of several coordinated solutions, all member states prefer coordinated to non-coordinated results and the probability of cooperation is high. In contrast, Rambo situations are a severe threat for regional integration; Rambos have no overall interests in cooperation at all, and they will not agree to majority rule or the delegation of decision-making competencies in order to solve deadlocks. Moreover, legalisation is unlikely to ensure the compliance of Rambos because the other member states lack the possibility of enforcing judicial rulings with decentralised sanctions. Only when the pay-off matrix of Rambos is changed by granting side payments or by negotiating large package deals can cooperation take place.