The Rarity of Rewarding

Explaining the drivers of compliance with international law has been a central concern of international lawyers (and many others) for decades. As international agreements have proliferated—exact numbers are elusive, but there may be 75,000 or more—understanding compliance has only grown in importance. In previous work we each have explored the question of compliance, focusing on the conceptual difficulties with compliance as a measure of effectiveness, the role of reputation, and other factors.¹ The literature on compliance with international law is rich, interdisciplinary, and, at this point, over three decades old. In Rewarding in International Law, Anne van Aaken and Betül Simsek seek to offer a new perspective on the enduring topic of compliance: that of rewarding.² In this essay, we suggest that rewards, when properly defined, are a conceptually interesting but empirically rare tool in the arsenal of states seeking to improve compliance with international law.

In Rewarding in International Law, van Aaken and Simsek shine a light on the role of what they term rewarding in international law, and they distinguish positive compliance incentives, which they term rewards, from negative compliance incentives. At one level this focus on the carrot rather than the stick is not new. Yet the extended treatment they offer of this conceptual frame strikes us as a salutary addition to the literature that can deepen our understanding of how compliance is achieved. Sanctions and other forms of negative incentives are common in international law; treaties may expressly contain sanctioning provisions, and even the law of state responsibility includes what are essentially sanctions as a way of remedying or responding to acts deemed internationally wrongful. Rewards, on the other hand, are less often discussed, and less often directly incorporated into international law. Nonetheless, an extended treatment of rewarding as a concept is novel and merits attention from international lawyers and others interested in understanding—and improving—compliance with international law.

In this short response to Rewarding, we advance two claims. First, we argue that one should conceptualize rewards and rewarding more narrowly than van Aaken and Simsek do. Second, we offer some preliminary thoughts about the frequency of rewarding in the international system, suggesting that it may be a rare phenomenon.

Refining Rewarding

Rewarding offers a typology of rewarding in international law. Disagreements about typologies are sometimes mere semantic nitpicks. In this case, however, we believe their typology introduces unnecessary complexity to an already robust existing literature. The term “rewarding” is helpful, but it should be narrowed substantially and used only to capture features of international law that are distinct from those already addressed in existing scholarship.

For example, in Part III of their article, van Aaken and Simsek distinguish “internal rewarding” from “external rewarding.” The former is defined as “the benefits of the agreed upon bargain.” An exchange of commitments in the form of a bargain lies at the foundation of every international agreement; characterizing the benefits of that bargain as a “reward” adds unnecessary jargon where none is needed. A trade agreement is an exchange of commitments and the parties each believe, ex ante, that they benefit from this bargain. Relabeling the bargain as an exchange of rewards does not advance our understanding.³

The term reward remains useful, but it should be reserved for what the authors refer to as “external rewards,” defined as rewards “‘on top’ of the cooperative benefit.”⁴ These rewards are, in other words, external to the agreement as a legal matter.

The authors also typologize rewards on another dimension—whether the reward is compensation for entering the treaty⁵ or for complying with it.⁶ Any benefit received for entering an agreement, whether formally reflected in a document or not, is simply part of the agreed bargain and nothing is gained by describing it as a reward.⁷

In our view, the real value of the rewarding framework is its consideration of benefits that are delivered after the agreement is entered into and that are not contemplated by the parties at the time of the agreement. In other words, rewards as a conceptual category should be limited to ex post and unanticipated benefits that are granted to a state to encourage compliance. With this more focused definition in mind, the insights of the paper are more easily understood and fit more neatly within existing scholarship.

Rewarding Compliance in the Face of an Incentive to Breach

The most straightforward example of rewarding occurs when one party to an agreement decides that it is no longer in its interest to comply. This could reflect some unanticipated change that alters the payoffs facing at least one of the parties.⁸ To illustrate, imagine a bilateral agreement under which states A and B agree to implement fishing restrictions to protect a shared stock of fish. After the agreement is in force, however, state A suffers a catastrophic crop failure, severe hardship, and widespread starvation.

Assume that both countries recognize that state A has a powerful incentive to address its food crisis by breaching the agreement and allowing unrestricted fishing in the relevant waters. The immediate crisis renders the benefits of preserving the fish stocks for the future insufficient to support compliance, and the agreement has too little “compliance-pull,” to use the late Tom Franck's phrase, to generate compliance.⁹

Prior to the crop failure, both parties preferred mutual compliance and protection of the fishing grounds—hence their assent to the agreement—and breach by either party was likely to trigger breach or countervailing measures by the other, leaving both parties worse off. After the unexpected crop failure, however, state A's incentives changed and its preferred policy choice is to breach, even if that triggers reciprocal non-compliance by state B. State B, on the other hand, continues to value preservation of the fish stocks.

What to do? State B could threaten A with some form of sanction. This could be reciprocal non-compliance, perhaps in the form of suspension or termination as contemplated by Article 60 of the Vienna Convention on the Law of Treaties, or it could go further and include additional consequences such as trade measures. If that threat is credible, and the resulting cost to A is high enough, compliance can be preserved.

Yet retaliation in the international arena has several drawbacks. First, it is costly to the retaliating party. Whether the strategy is termination of the treaty, trade sanctions, diplomatic sanctions, or even military action, both parties pay a price. Second, and related to the first, the threat to retaliate may lack credibility. The threat of retaliation is intended to induce compliance. Once the breach occurs, there may be little reason to go through with the threat, undermining its effectiveness to begin with.¹⁰

State B has another option. It can offer to reward state A for compliance. It might, for example, provide food aid in sufficient quantity to persuade state A to comply until the crisis has passed.

In this hypothetical, we refer to state B's action as a reward because it is an ex post and unanticipated benefit. Providing the reward is worthwhile to B as long as A's continued compliance with the agreement is worth more to B than the cost of the reward. Although the reward alters the terms of the bargain (at least for some period of time), the changed circumstances make it an efficient outcome.

Challenges to Rewarding

We see at least four reasons why rewarding is likely to be rare as an empirical and, perhaps, prescriptive matter: informational challenges, moral hazard, cost, and domestic political impacts. We briefly address each of these in turn.

If rewards are available, every state has an incentive to seek these benefits from their partners. In a world of complete information, unjustified demands for rewards will typically be ignored because the putative rewarding party will know the reward is not necessary to induce compliance.¹¹ With asymmetric information, however, it becomes more difficult for a would-be rewarding country to decide how to respond to a request for a reward. There is a danger that any unexpected event will trigger opportunistic demands for rewards and threats of breach. Rewarding countries will fear that they may be providing a reward that is unnecessary—that they may simply have been fooled or bluffed. Similarly, every time a reward is refused, there is a risk that the rewarding country has misjudged and the threat to breach was not a bluff. The risk of both types of errors makes the expected benefits of rewarding smaller, and so reduces its appeal.

The possibility of a reward also creates a moral hazard problem.¹² Rewarding operates like insurance. When compliance becomes too expensive to be worthwhile for a state, its partner steps forward and offers a reward to make compliance more attractive. If that insurance is available, the insured state can take greater risks, knowing that the downside risk will be at least partially offset by the reward, which operates like an insurance payment. Consider our earlier example of a country facing a crop failure. If rewarding is common and the country can rely on food aid being provided in the event of a crop failure, it has less incentive to secure emergency stores of food or undertake policies to encourage sustainable farming. The rewarding state, of course, does not want to underwrite risky behavior by a treaty partner, so it may have an incentive to develop a reputation for refusing to reward its partners and, instead, simply allow the breach to happen.

Cost presents a third possible reason why rewards appear rare. Sticks benefit from a crucial feature that carrots lack. A threat of retaliation has the potential to promote compliance for free. (Almost: van Aaken and Simsek rightly note that even threats can be costly, not only in political terms, but literally, in that threatening trade sanctions, for instance, can be costly by inducing distorting adaptations in local and foreign firms.) As we note above, actual retaliation, as distinct from reciprocal withdrawal of compliance, is almost always costly for both parties. When a threat works, however, there is no need to ever impose the retaliation. This makes it an attractive tool for a state attempting to encourage its partner to comply. A reward, on the other hand, works only when it is actually delivered. A reward will always be costly to the rewarding state because it must be provided to be effective.¹³

Finally, rewards may, and often will, have negative domestic political effects. Sanctioning (or even attacking) other states is often politically popular at home because it provides domestic actors with a common adversary. Rewarding, on the other hand, invites domestic political opponents to accuse decision-makers of weakness and an inability to hold other nations to their bargain. A reward is only needed, after all, if the foreign country is threatening to breach an existing and agreed upon legal obligation. And since a reward requires resources, some domestic constituency must be asked to pay for it, ensuring at least some degree of local opposition. In short, rewarding a country in response to a threat to breach is unlikely to be politically popular.

Van Aaken and Simsek themselves add an additional, important, limitation on rewards: what they term, in Part IV, the “rewarders’ dilemma.” In a multilateral framework, who provides the reward? The reward is in a sense a public good as well as a private good. It is private in that it is “consumed” by a single party but public in that it can benefit, via its change in behavior, all the parties. If our fisheries example is extended to a multilateral treaty, one can readily see this dynamic. All states gain when state A is rewarded to prevent breach. But which of the other states will step up to provide the reward? What will stop some from free-riding? This issue can, in theory, be addressed ex ante via a mechanism in the treaty framework for sharing the burden of rewards, but in practice, such mechanisms are very rare. This suggests that, at least at the stage of entering into the agreement, the parties do not plan on making extensive use of rewarding.

For all of these reasons, we believe that rewarding as we have used the term—ex post benefits conferred, beyond those contemplated by the agreement, for the purpose of encouraging compliance—is likely to be and remain a rare phenomenon.

Ignoring

If rewards are, indeed, rare, what do we imagine happens in cases like the above hypothetical involving a crop failure? Rewarding focuses on a binary choice between carrots and sticks. Yet parties actually face a tripartite choice when a counterparty breaches: reward, sanction, or ignore. The rewarding dynamic rightly highlights that the existing debates over compliance with international law have tended to over-emphasize negative incentives—sanctions and countermeasures and treaty exits—and not inducements. This an important point. In the same vein, ignoring should not be ignored.

As was demonstrated long ago in the relational contracting literature associated with Stewart Macaulay, or the “order without law” literature associated with Robert Ellickson, parties in the real world do not typically seek redress through coercive means but rather try to work things out in a manner that may, at times, simply allow the breaching party to go about its business.¹⁴ If this is often true in the domestic realm, where well-functioning hierarchical courts allow contracting parties to readily enforce their terms, we believe it is likely to be even more true in the international realm. Just as rewarding was due for a deeper exploration, one helpfully provided by van Aaken and Simsek in their recent article, perhaps too ignoring ought to be systematically analyzed for its role in international law.

Conclusion

Carrots and sticks are commonplace terms—and commonplace tools—in diplomacy generally. In international law, however, sticks predominate. Many provisions in international treaties exist that aim to discipline the process of sanctioning, such as the World Trade Organization dispute settlement understanding. By comparison there are far fewer (perhaps no?) examples of legal provisions aimed at regulating the rewards process. This does not mean rewards cannot be a useful approach in promoting compliance, but in a world of tens of thousands of international agreements with a wide variety of provisions and approaches, the dichotomy is striking.¹⁵ We believe that while interesting conceptually and perhaps useful in niche cases, rewarding has never been and is not likely to become a widespread approach to enhancing compliance with international law.