Designing Rewards

This essay engages with the literature on regime design and problem structure in order to propose a set of questions to frame our thinking about rewarding in international law. I am particularly interested in finding out whether the nature of an objectionable policy or behavior should be taken into account when designing the reward. There is a vast literature in International Relations that investigates State behavior with the goal of unpacking the inherent challenges for enforcement and compliance. This is in light of the absence of effective enforcement mechanisms in International Law. Thus, in order to understand shortcomings in State compliance, scholars have increasingly looked into the challenges for behavior modification. According to rationalists, the greater these challenges, the less likely we are to observe timely compliance. Managerialists, on the other hand, point to the need to provide technical capacity and catered exceptions in order to obtain compliance. Of course, this literature is prolific with respect to punishment mechanisms and is comprehensively analyzed by Anne van Aaken and Betül Simsek (2021). But how can insights from this literature help us better understand the relationship between rewards and the objectionable policy or behavior? How can a deeper understanding of the objectionable policy or behavior help us design rewards that will increase the chances of State compliance?

Scholars within the regime design and problem structure traditions have long thought systematically about how the nature of issues (problems) is associated with, or even influences, challenges for compliance. For example, when State compliance entails the provision of a public good, challenges to collective action come into place. As a result, full compliance may be halted by the ’free rider‘ problem. Coastal fishing illustrates the point. When two States share a coastal body of water where fishing represents an important source of income for local communities, efforts to prevent overfishing constitute a collective action problem. To borrow from Elinor Ostrom (2015), fish stocks can be analyzed as a common pool resource where exclusion is difficult and consumption is rival. Simply put, States cannot regulate effectively without cooperation, but given the nature of the problem, there are incentives to free ride on another’s efforts to prevent overfishing.

This essay highlights the role of institutional architecture with respect to the punishment versus reward dichotomy and proposes a fresh look at the role of problem structure. Is the structure of a problem or issue an important element, when it comes to understanding the dynamics of rewarding as a compliance-enhancing mechanism? I explore this question on the shoulders of scholars who have studied institutional design and choice architecture within both rationalist and a behavioral science frame.

Incentives, Capacities, Information and Norms

Political scientists and international relations scholars have long recognized the characteristics of an issue (or problem) as an important aspect of the compliance equation. From a rationalist point of view, even if a Head of State has incentives to revert the objectionable policy, compliance may be hindered by obstacles that are beyond his or her control. Take the example of government repression. A government that is incentivized to end to repression – for example, via contingent preferential trade agreements – may not be able to fulfill its commitment because it takes time to swap deeply rooted repressive practices (Christian Davenport 2004). The 2001 special issue of the journal International Organization undertook a systematic investigation of the nature of a set of ubiquitous characteristics and their response to several omnipresent regime design features. By design features, the authors refer to institution-building blocks such as: How many States will be invited to participate in the agreement? Will there be an institutionalized bureaucracy to oversee implementation of the agreement? Is the agreement limited in time? Can States derogate from the agreement? What are the decision-making rules within the agreement? It is easy to see that the universe of international agreements recorded thus far reveals great variation in these design features. But how can institutional design be mobilized to promote State compliance? The answer to this question presupposes a thorough understanding of the nature of the problem at hand.

While Ronald Mitchell (2006) emphasizes the role of incentives, – for instance those inherent to the strategic game being played, the State capacity to comply with commitments, the availability as well as capability to process information, and the existence of binding norms, – others before him have proposed a distinct framework. For Barbara Koremenos, Charles Lipson and Duncan Snidal (2001) the key characteristics of a problem revolve around: 1) distribution; 2) enforcement; 3) number; and 4) uncertainty. Distribution, much like “incentives” for Ronald Mitchell, refers to the dynamic amongst the actors that are relevant to a particular situation. For example, when the issue is a dispute over territory, the States involved in the dispute are likely to perceive the problem as a “zero-sum game.” Resolving the dispute may entail granting sovereignty over the disputed territory to one State to the detriment of the other. Zero-sum games are typically protracted disputes which demand a special set of incentives and bargaining strategies in order to be overcome.

A second important characteristic, according to Koremenos et al. is enforcement. Enforcement is associated with the challenges of monitoring compliance as well as other characteristics that may impact the threshold for compliance-consistent incentives. To recapture the examples above, it is probably easier to monitor compliance with an agreement that resolves a territorial dispute than to verify that an authoritarian regime has ceased to mobilize repression. The regime design literature argues that these differences are inherent to problem structure; it also urges policymakers as well as scholars to take these characteristics to heart.

Finally, Koremenos et al. propose that number is an issue characteristic that reveals the set of actors whose preferences and behavior are relevant to the analysis of a situation. They also invite us to vet for uncertainty, or the degree to which relevant actors can predict future circumstances; these circumstances range from the behavior of other relevant actors (uncertainty about behavior), to the context within which the behavior takes place (uncertainty about the world), to the probability of changes in preference (uncertainty about preferences). Underlying these efforts is a commitment to fully grasp the characteristics of an issue in order to design institutional mechanisms that are Pareto-enhancing. Conversely, when we analyze the effectiveness of a particular set of institutions, for instance rewards and punishment mechanisms, it is essential to take into consideration the underlying structure of the problem that these mechanisms attempt to mitigate.

Unpacking the Baseline through Problem Structure

The concept of the baseline, i.e. the set of conditions that form the context against which punishment and reward operate, is a key feature of van Aaken and Simek’s theoretical argument. The authors make a convincing case for the importance of analyzing the baseline on a case-by-case basis, in order to design policies that can be genuine rewards. I propose that an exploration of problem structure may be a pathway to assessing the baseline. In other words, what are the distributional consequences of the objectionable policy? What are the enforcement challenges associated with the objectionable policy? How many relevant actors (number) are there whose preferences ought to be taken into consideration with respect to the objectionable policy? Is there uncertainty related to behavior, the state of the world, or preferences that may impact decisions with respect to the objectionable policy? The answers to these questions will help to paint a comprehensive picture of the problem and may provide a stronger case for the superiority of reward over punishment.

Designing and Calibrating Rewards in Response to Problem Structure: the Role of Regime Design

It follows that the design and calibration of rewards can benefit from an understanding of the distributional and enforcement challenges that surround the objectionable policy. It can also capitalize on the actual mapping of relevant actors, the degree as well as the type of uncertainty that the problem embeds. For example, if a reward is directed at the resolution of a territorial dispute, where States perceive the problem as a “zero-sum game” almost by definition, the State that makes the most concessions might be entitled to a bigger reward, as per each State’s perception given its baseline. Conversely, if the reward is directed at the resolution of a problem where enforcement is not trivial, as in most objectionable policies related to State repression, the reward may be split into two interrelated measures: compliance per se and monitoring; in other words, the State would be rewarded for eliminating repression and also for enabling independent monitoring.

Van Aaaken and Simsek point to the tenuous line that separates reward and punishment; they already stress the importance of a case-by-case analysis in order to identify the space for relative gains that a reward policy presupposes. This essay expands on this point to propose a more direct dialogue with literature on regime design and problem structure. In this sense, such a dialogue could enable further theorizing on the relationship between institutional design features and the design of reward policies. Underlying these propositions, there is a potential for a fresh, empirically informed, research agenda. Here, the influential work of Barbara Koremenos (2016) could be further explored in order to analyze the role of regime design elements – membership, scope, centralization, control and flexibility – for the conception of effective reward policies. These design elements were conceived as dependent variables, derived from a direct observation of State behavior and international lawmaking. For example, control refers to the decision-making prerogatives within a reward policy: who designs, who negotiates, who monitors, who enforces. Conversely, flexibility refers to the possibility of derogation as well as the incorporation of temporal elements, such as duration or withdrawal. In the context of our thinking on rewarding in international law, these design elements could be mobilized to mitigate some of the problem characteristics identified earlier. It is often the case that flexibility elements, such as duration and temporary derogation, can contribute to mitigating the unanticipated adverse consequences of uncertainty. Mobilizing flexibility elements may enhance the potential for success of a rewarding policy.

Design elements and behavioral economics

A final word to reinforce the argument presented by van Aaken and Simsek: when they argue that the superiority of reward over punishment passes the test of both the rationalist and behavioral approach, the authors make a compelling argument. Nevertheless, within behavioral economics approaches, the role of regime design elements remains underexplored. Perhaps there is an opportunity to analyze the architecture of reward policies in light of findings within the fields of choice architecture and regime design. The work of Richard H. Thaler and Cass R. Sunstein (2008) offers several possibilities. And there is always the intersection between choice architecture and rationalist approaches which have advanced our knowledge with respect to institutional design in areas where risk aversion prevails. Here, the role of flexibility elements has been thoroughly explored. What inferences can we draw from this literature for the conception and implementation of effective reward policies? More specifically, how can flexibility elements such as duration and withdrawal clauses be mobilized in order to optimize reward policies?

With this essay, I hope to have proposed a set of questions that warrant a dedicated investigation of the contribution that the regime design literature can bring to the analysis of rewarding as an effective mechanism of international policymaking.