PDF Print Version                                                   Vol 2, Issue 10
Editorial board: Anne van Aaken, Jutta Brunnée, Jan Klabbers, André Nollkaemper (editor-in-chief)

Anne van Aaken
University of St. Gallen

Behavioral International Law and Economics

 

 

Public international law scholarship opens up ever more to social science in order to answer questions of treaty design and legal interpretation; social science can make international lawyers better lawyers, to quote Anne-Marie Slaughter. The promise of those ventures is to be able to better understand “how international works” (or not) in order to guide treaty design as well as international adjudication. Whereas the rational choice approach to international law has been widely accepted in legal scholarship and international relations (IR) theory, behavioral challenges to the rational choice paradigm in international law have hitherto not been systematically explored.[1] Nevertheless, behavioral law and economics has been successfully applied to national law and is now used in public policy (in the US, the UK and by the European Commission) but not in international law since law not only automatically acts as a “choice architecture” (Sunstein/Thaler), but can also function as a debiasing mechanism. Behavioral economic and psychological insights have furthermore been used by IR scholarship under the heading of political psychology. However, the latter research does not take into account international norms and is mostly confined to individual actors (elite decision-makers) and security constellations. It is therefore a promising research agenda to build on all those insights and explore the benefits and challenges of extending the behavioral law and economics approach to public international law: behavioral international law and economics (BIntLE). With a view to ensuring the “unity of knowledge” (E.O. Wilson), it seems time to draw on those different insights to further refine our understanding of international law.

This Reflection will sketch out the three pillars on which the BIntLE research rests (I.), discusses some methodological challenges (II.) and gives some examples in treaty design and adjudication where BIntLE has the potential to either confirm or challenge rational choice insights (III.).

I. The Three Pillars of BIntLE

The standard models of rational choice (RC) theory, using instrumental rationality, have minimal assumptions about human cognition as well as the human preference function. Following the traditional international law assumptions that have prevailed ever since the Westphalian peace, the nation-state has mostly been analyzed as a unitary actor, or what has been described as a “black box” or “billiard ball”, assuming rational behavior. RC analysis has been used to diagnose substantive problems and frame legal solutions and explain or re-conceptualize the structure or function of particular international legal rules or institutions. Many tools of economic analysis are used in the endeavor: game theory, contract theory, price theory, principal-agent-theory and collective action or social choice theory and political economy analysis as well as the notions of externalities, public goods and commons.[2] These assumptions were transferred to states and the assumed preferences were power or military might (realist) or utility and welfare (institutionalist) by IR scholarship using RC and the same tools as economists.

However, based on many experiments, the rationality assumption has been called into question by cognitive psychologists and behavioral economists. Based on prospect theory,[3] the research experimentally confirmed that individuals are systematically only boundedly rational, that is, that their cognitive abilities sometimes lead to objectively wrong decisions (e.g. in probability calculus) and that their decisions depend mostly on the context and the frame in which a decision takes place. They commit fundamental attribution errors and are subject to an endowment effect, that is, objects in their possession are valued higher than those which are not. Individuals are loss averse, i.e. they have an asymmetrical attitude towards gains and losses: their utility is less increased by gains than by averted losses and they are risk averse in a loss frame. What is viewed as a loss or a gain depends on the reference point chosen (often set by law or entitlements).[4] Individuals also succumb to the status quo bias, which makes default options (also in the law) crucial since actors stay with it. Opting out of the default is much rarer than opting in: choice architecture therefore matters.[5] Although the so-called self-serving bias is well known, individuals have other-regarding preferences, inequity aversions and fairness preferences but can are also be more spiteful. They are willing to incur costs to punish free-riders or unfair behavior, especially if they attribute norm-deviating behavior to the “type”, rather than to the situation (reputation is therefore a relational concept).[6] Altruistic cooperation is an important behavioral force and might explain international cooperation where RC would not expect it. These insights are also used by political psychology in IR theory, with a special view on security and crisis situations, analyzing the behavior of head of states or other decision-makers, e.g. the behavior of President Kennedy in the Cuban Missile Crisis.[7] Bringing those research strands together is the aim of BIntLE.

II. Methodological Challenges

States cannot be put into the lab. Thus the question arises whether insights from individual psychology can tel quel be transposed to a corporate actor, such as a state. RC does not raise the same methodological problem since it is an assumption, not a descriptive theory of actual behavior. It all depends on the relevant units of analysis. If those are individuals, like international judges or arbitrators, behavioral research can be more applied easily (although panel effects might be present). Elite decision makers, like treaty negotiators or heads of state can be attributed to the state and may count as state behavior (as in Art. 7 VCLT) but there might be 1) a principal-agent problem (agents exhibit e.g. less endowment effect than principals, since they decide on other people´s possessions) and 2) a group decision situation, since even dictators rarely decide without group advice. If the unit of analysis is the state as such, BIntLE gets complex but is more aligned with approaches in IR that look at internal political processes (as e.g. in the theory of the norm-spiral in human rights law).[8] Furthermore, if Slaughter is correct that the international world starts to look like a “Lego World”,[9] the different component parts of the state can be looked at separately. Still, the aggregation problem can depend on municipal institutions and might dif
fer from authoritarian regimes to parliamentary systems. Another aggregation mechanism is the market and through the market, international human rights or environmental norms can be effectuated under certain circumstances and if consumers have a preference for fairness (e.g. not buying products fabricated under human rights violations) or for environmental protection (sustainable foresting or fisheries).[10]

III. Applications

There are many potential applications in general and specific international law. I will pick just a few on treaty design and international adjudication.[11] The question of why states conclude treaties and under what circumstances has been extensively analyzed by rational choice scholars using economic contract theory: states enter treaties whenever they assume that the gains from cooperation are higher than from non-cooperating. Guzman[12] argues from a RC perspective that consent is difficult to reach and therefore more forms of nonconsensual law-making should be used. BIntLE may identify additional arguments for the reluctance to conclude treaties. Korobkin/Guthrie[13] find that some common heuristics (i.a. anchoring, availability, self-serving evaluations, framing, the status quo bias) are likely to influence the negotiator’s decision-making processes in bargaining, leading to less consent. Also, linking treaties seems to be less beneficial for agreement than supposed by RC since parties tend to overestimate the value of their own concessions and undervalue those of their adversary; a reluctance to accept losses on any dimension of linked agreements explains bargaining impasse (e.g. in the Doha Round). Disentangling which biases might either inhibit beneficial treaty conclusion or make states enter treaties which are not beneficial sheds insights on how negotiations can be influenced.

Since complete contracts are impossible to draft, economic contract theory finds that overly strict and inflexible contracts may impair the joint surplus of the contracting parties ex post because they do not accommodate unforeseen circumstances.[14] A trade-off arises between flexibility and commitment. Behavioral research suggests a more nuanced view, arguing “that a contract provides a reference point for the parties’ trading relationship: more precisely for their feelings of entitlement.”[15] Flexible contracts are even more permissible in regard to licit behavior and permit so-called shading[16] in ex post performance, while under rigid contracts much less shading can occur. Shading will occur if parties to the contract interpret ‘their’ reference point with a self-serving bias or deem contracts unfair. Although parties do not feel entitled to outcomes outside the contract, they may feel entitled to different outcomes within the contract. If a party does not get what it feels entitled to, it is aggrieved and shades by providing perfunctory rather than consummate performance, causing deadweight losses.[17] Thus, whereas flexible contracts dominate rigid ones under RC assumptions, this is not necessarily the case with behavioral assumptions. This insight calls for fewer usage of indeterminate legal terms and instead more explicit flexibility mechanisms in treaties.[18]

Reservations of treaties can be reconsidered under the default option insights.[19] Galbraith finds that depending on how the reservation option is framed, there are statistically significant differences in how states react: on average, where states have the implied authority to reserve out of ICJ jurisdiction, 95% continue to accept it, but where states can explicitly opt into ICJ jurisdiction, only a mere 5% of state parties do so. One may use this insight for objections to reservations and conjecture that the opt-in rule under Art. 20 (5) VCLT leads to far fewer objections especially in human rights treaties, where reciprocity plays no role. Thus, BIntLE submits an additional argument for changing the default option of interpreting the omission of an objection as an implicit consent to another rule: either the omission is interpreted as an objection (opt-out rule) and/or a minority of states’ objections is enough to invalidate the reservation (reversed Art. 20 (2) CERD model). This would enhance the integrity of treaties and solve the problem of decentralized interpretation of Art. 19 (c) VCLT.

Another promising venue is research on judges and arbitrators. National judges have been extensively analyzed form a RC perspective as well as a behavioral perspective, using experiments.[20] Judges often, but not always, rely upon intuition to process information and make decisions: irrelevant numerical anchors affect damage awards; framing of disputes impacts judges’ evaluations; judges overestimate their skills in assessing witness credibility as well as facilitating settlement and avoiding the influence of cognitive biases. This research provides a baseline for testing the cognition of international law adjudicators. Although there are – as of now – no experiments with international judges or arbitrators, one may reasonably expect that they show the same biases and heuristics as domestic judges.[21] Procedural provisions as well as group decisions might mitigate those effects; one reason why sole arbitrators, e.g., are not a good idea.

IV. Conclusion

In spite of methodological challenges, the proof of the pudding is in the eating: the success of this approach will be measured by its applications and its usefulness for the design and interpretation of international law. To use a coin minted by Einstein: “Everything should be as simple as it can be, but not simpler.” I put the burden of proof on BIntLE, but I submit that RC is sometimes too simple to explain international law phenomena and might therefore also lead to faulty treaty design. Thus, the analysis has to be done step by step, analyzing different fields of general and special international law, taking into account empirical findings.